US 5,043,807 AGrant
Three Dimensional Composite Video Motion Detection
Issue Date:1991-08-27
•63 Claims
•11 Drawing Sheets
Abstract
A system and method for motion detection in a composite video signal is disclosed. These dimensional (horizontal, vertical and time) analysis is performed on actual picture element content to monitor variations therein. Vector generation, approximation and analysis gives information on picture element content for intra-field, inter-frame and inter-field signal processing. Additionally, noise immunity through selection of a convergence zone for the disclosed vector analysis is also provided. Control signals for further video signal processing are likewise generated.
Metadata
Assignee
- Zenith Electronics Corporation
Inventor
- Khosro M. Rabii
Application Information
Application Number:US 3560384
Filing Date:1989-05-23
Priority Date:1989-05-23
Art Unit:262
Classifications
IPC:
H04N 718H04N 5213H04N 701H04N 978
Field of Search:
358105;11;167;36;140;31
Patent Drawings (11 sheets)
Description
Background of the Invention
The present invention relates generally to television systems. More particularly, the present invention relates to improved definition television systems which seek to provide a higher quality picture from now-standard broadcast signals. Even more particularly, the present invention seeks to provide improved video signal processing and noise immunity characteristics in response to signal content to provide desired improvements in picture display.
High definition or enhanced definition television ("HDTV") systems seek to provide a higher quality television picture by improving the characteristics of the broadcast or transmitted signal. Several such systems are now known; all, to date, require additional electromagnetic spectrum bandwidth to provide the enhanced broadcast signals. In contrast, improved definition television systems seek to provide a higher quality television picture by improving the characteristics of the signal processing at the television receiver. Improved definition systems seek to give a better picture within present broadcast standards, e.g., the six megahertz (6 mhz.) per channel standard in the NTSC system.
Standard video signal processing uses some type of filtering to separate or comb the 3.58 megahertz color carrier signal from the 4.25 megahertz bandwidth luma signal. Desirably, the implemented filtering prevents artifacts or distortions in the displayed picture. Without some filtering, the amplitude modulated color signal can produce a grid appearance on the screen in response to chroma signals that are mistakenly recognized as and then processed as luma signals.
However, several known filtering techniques each can introduce undesired artifacts in the displayed picture. For example, the simplest filtering technique, notch filtering, can undesirably reduce resolution. A notch filter removes video signal in a notch surrounding the 3.58 megahertz chroma carrier. Alternately, a notch filter provides luma in the bandwidth of zero to about 3.1 mhz. However, that also removes luma within the same region or above the filter cutoff frequency, thereby reducing horizontal resolution. This becomes readily apparent when the television camera pans back from a structure, such as a tall office building, comprised of many smaller, horizontal elements, such as windows; at some point the building and windows will become blurred--that point occurring when the level of detail necessary to resolve the window edges passes into the notch above about 3.1 MHz.
Alternate filtering techniques include line filters. Such filters rely on the known characteristic R of standard video signals that alternate lines of each picture have opposite polarity chroma signals. Pure luma may be extracted by adding the video signals for adjacent lines (typically stored in a line buffer or memory), thereby cancelling chroma. However, this does not always work; for example, when one line has no chroma, simply adding it to the next line will not cancel the chroma. This creates a loss of vertical resolution and gives the appearance of "hanging dots" on the screen. The hanging dots appear in a zipper pattern which seems to move back and forth on the screen as the non-cancelled chroma information is processed as luma.
One other method is known as frame filtering or combing. This technique relies on another known characteristic of standard video signals, namely that the same line in alternate frames carries chroma of opposite phase. However, this technique can only be used when succeeding frames have identical video content. If not, then the displayed picture is undesirably distorted.
Of course, any filtering technique must completely separate the chroma information for chroma processing. In the worst case, an inadequate filter can result in luma being interpreted as chroma. This creates highly undesirable cross chrominance in the displayed picture. Likewise, noise in the television signal can undesirably affect the processing of the luma and chroma components.
Each of the known and described filtering techniques provides adequate performance and an acceptable display in some circumstances. However, each suffers drawbacks or inadequacies as described in the foregoing examples. Accordingly, the present invention seeks to provide an improved signal processing technique that generally overcomes the deficiencies of the prior art and provides improved picture display. The present invention lies, in part and fundamentally, upon recognizing actual picture content13 within selected noise levels.
A further object of the present invention lies in the implementation of techniques for choosing one or more processing methods in response to received video signals.
A further object of the present invention lies in providing an improved picture quality based upon actual picture content.
A further object of the present invention is to provide enhanced video signal processing through recognition of selected characteristics in the video content of received picture signals.
Another object of the present invention lies in providing improved displayed picture quality through reduced noise susceptibility in chroma/luma processing. A broader related object lies in improving noise immunity in composite video processing.
Yet a further object of the present invention is to provide a motion detection technique for improved signal processing to provide an improved picture display.
Brief Description of the Figures
Those features of the present invention which currently are believed to be novel are set forth with particularity in the appended claims. The foregoing objects, in addition to other further objects and advantages, of the present invention may be more fully understood and appreciated by reference to the following detailed description taken in conjunction with the accompanying drawings of which:
FIG. 1 illustrates a first chroma/luma filtering method for separating chroma and luma in a composite video signal;
FIG. 2 represents a second method, line filtering, for separating luma and chroma in a composite video signal;
FIG. 3 illustrates a third method, frame filtering, for separating luma and chroma information in a composite video signal;
FIG. 4 illustrates an exemplary chroma signal;
FIG. 5 illustrates a dot matrix, pixel pattern for use in a motion detection technique of a preferred embodiment of the present invention;
FIG. 6 provides a further diagrammatic representation of a pixel matrix for three fields related to the motion detection technique of a preferred embodiment of the present invention;
FIG. 7 illustrates the relationship between the four times chroma (4 Fc) sampling signal and the fundamental chroma signal;
FIG. 8 provides a further diagrammatical representation of pixel information for use in the motion detection technique of a preferred embodiment of the present invention;
FIG. 8A illustrates the horizontal, vertical and time axis for the pixel diagram of FIG. 8;
FIG. 9 represents a first set of vectors for use in a motion detection technique of a preferred embodiment of the present invention;
FIG. 10 illustrates a second set of vectors for use in a motion detection technique of a preferred embodiment of the present invention;
FIG. 11 illustrates a schematic representation of a combination of variables for estimating motion variables in accordance with the technique of a preferred embodiment of the present invention;
FIG. 12 illustrates additional signal combination schematics for providing motion variable information and prediction in accordance with a preferred embodiment of the present invention;
FIG. 13 illustrates a scaling function for use in processing motion variables according to a preferred embodiment of the present invention;
FIG. 14 illustrates a combination logic schematic for providing output signals indicative a desired filtering technique in response to the motion variables generated in accordance with a preferred embodiment of the present invention;
FIG. 15 illustrates a general digital television system envisioned as the operating environment for the preferred embodiment of the present invention;
FIG. 16 illustrates a combinational logic implementation for effecting a desired filtering technique in response to the motion detection signals; and
FIG. 17 illustrates various logic state combinations for implementation of multiplexing techniques used in the logic schematic of FIG. 15.
Detailed Description of the Preferred Embodiment
Generally, the present invention provides an improved technique for providing display of received broadcast television signals. Initially, a preferred embodiment analyses actual picture content in two or more display lines in a current display field and corresponding display lines in two or more previous display fields. Vectors based on this analysis are then processed to provide an indication of motion in the picture within an acceptable range governed by use selected noise level parameters. The technique of the present invention then selects one of three or more techniques for separating luma and chroma from composite video. The composite video signal and picture content is analyzed; selected comparisons are made to provide information that reflects actual picture content. That information is employed to select one of the luma/chroma separation techniques.
Referring now to the drawings, FIG. 1 illustrates a first technique that employs conventional one dimensional filtering methods where luma is derived from a low pass filter and chroma from a band-pass filter. More particularly, a band pass filter 20 and a low pass filter 22 both receive a composite video signal on an input line 24. The band pass of filter 20 is centered at 3.58 MHz (typically .+-.0.5 MHz), thereby providing separated chroma as an output. The low pass filter 22 has a cutoff frequency desirably at about 3.0 MHz, thereby providing separated luma as an output. Alternately, the low pass filter 22 could be replaced with a notch filter arrangement that blocks only those frequency components of the received composite video signal that are passed by the band pass filter 20.
FIG. 2 illustrates a second technique for chroma/luma separation which employs a two-dimensional filter (line comb). In the system of FIG. 2, a one-line delay 26, labeled Z.sup.-H receives a composite video signal on a line 28. A first summer 30 and a second summer 32 also receive the composite video signal from the line 28. The first summer 30 adds the received composite video signal with the output of the one-line delay circuit 26. The second summer 32 subtracts the output of the one-line delay circuit 26 from the receive composite video signal. It will be appreciated that, because the chroma signal is opposite in phase for adjacent lines, at any one time, the output of the one-line delay circuit 26 will represent a composite video signal in which the chroma signal is exactly opposite in phase to the input signal on the line 28. This may be represented illustratively by designating the received composite video signal as (L+C) and the output of the one-line delay circuit 26 as (L-C).
Thus, the output of the summer 30, at which the received composite video signal from line 28 and the output of the one-line delay circuit 26 are each positively combined, represents a video signal having only luma, although at twice the normal magnitude, on a line 34. The signal on the line 34 may then be scaled at a scaling circuit 36 to provide luma on an output line 38. Correspondingly, the second summer 32 receives the composite video input signal from the line 28 and subtracts from it the output of the one-line delay circuit 26. In this fashion, the output on a line 40 represents only chroma information, at twice the normal magnitude. The signal may then be scaled at a scaling circuit 42 to provide a chroma signal on an output line 44. In this fashion, separate luma and chroma signals are provided.
Referring now to FIG. 3, therein is shown a system for implementing a third technique, known as a three dimensional filter or frame comb, for separating luma and chroma in a composite video signal. In this technique, a one frame delay 46, illustrated as Z.sup.-525H, receives a composite video signal from an input line 48. The output of the one frame delay 46 at a line 50 thus represents composite video exactly one frame delayed from that received on the input line 48. This system takes advantage of the known characteristic that alternate frames of a picture carry chrominance having opposite phases. Thus, the input and output to the one frame delay 46 in FIG. 3 may be represented similarly to that of the one line delay 26 of FIG. 2. That is, the input may be represented as (L+C), and the output may be represented as (L-C).
The output of the one frame delay 46 is supplied as a positive input to a first summer 52. The summer 52 also receives, as a positive input, the composite video signal from the line 48. Because of the opposite polarity of chroma in these two signals, the output on a line 54 thus represents a pure luma signal at twice normal magnitude. A scaling circuit 56 reduces that to a signal having pure luminance at the received magnitude. Correspondingly, a second summer 58 receives the output from the one frame delay 46 as a negative input and the composite video signal from the line 48 as a positive input. The difference signal on a line 60 then represents pure chroma at twice normal magnitude. A scaling circuit 62 scales that to provide pure chroma signal at normal magnitude as an output.
As mentioned, the line comb operates on the principle that the chroma phase between two successive lines in a field are opposite, and the frame comb operates on the principle that the chroma phase of corresponding lines a frame apart are opposite. FIG. 4 generally illustrates chroma waveforms showing the characteristics utilized in the line and frame combs of FIGS. 2 and 3, respectively. It should be appreciated that the chroma subcarrier frequency (F.sub.C) is an odd multiple of the horizontal scanning frequency (F.sub.H) divided by 2, that is 455/2, then (455/2) F.sub.H =F.sub.C and 525 F.sub.frame =F.sub.H. FIG. 4 illustrates a first waveform 70 which represents a chroma signal for a first line. As shown, two and one half cycles of F.sub.C equals one line (1H). FIG. 4 illustrates a chroma signal for the succeeding line in the picture as waveform 72 (in dashed lines). Again, a line equals two and one half cycles of F.sub.C. As shown, also, the chroma for the second succeeding line is opposite in phase to that for the first line. The second line also corresponds to the chroma for the first line in the next frame, i.e. line 526 from the first frame. It is this relationship that the filtering techniques of FIGS. 2 and 3, line and frame comb, take advantage of to provide the desired luma/chroma separation.
Thus, for the line comb, two successive lines can be represented by luma (L) plus chroma (C) and luma (L) minus chroma (C) or (L+C) and (L-C). Luma is derived by adding the components (normalized to one-half (0.5)) and chroma is derived by subtracting the components (normalized to one half (0.5)). The frame comb operates in a similar manner except that corresponding frames are processed instead of lines.
All three luma/chroma separation techniques have their advantages and disadvantages. In the one-dimensional filtering method of FIG. 1, high frequency components of luma information are lost in the luma low pass filter 22. In the two-dimensional filter (line comb) of FIG. 2, dot patterns (hanging dots) are produced when chroma changes. The three dimensional frame comb of FIG. 3 produces objectionable results when there is temporal variation in the image. The invention, as exemplified by the preferred embodiment, is therefore directed to a technique for selecting which luma/chroma separation method to use at any given time. The preferred embodiment of the present invention generates and responds to unique criteria for separating luma from chroma for each pixel in the picture.
In the preferred embodiment of the present invention, pixel elements from each of three succeeding fields are analyzed to recognize selected characteristics in the displayed picture. Recognition of particular characteristics provides an indication of which of the three techniques for luma chroma separation to use for each pixel element in the displayed picture.
Referring now to FIG. 5, therein is shown three successive fields of video, each field including a plurality of horizontal scanning lines comprising nine hundred ten (910) pixel samples. It will be appreciated that field I represents the odd lines from the first frame. Field II then follows in time after field I and is interlaced therewith by providing the even lines from that same frame. Field III then follows field II and is overwritten over the odd display lines from field I so that it too is interlaced with field II in the displayed picture.
Referring now to FIG. 6, therein is shown a vertical slice of the three fields illustrated in FIG. 5. The diagram of FIG. 6 shows the relationship of the chroma phase for the various pixels in the succeeding fields and succeeding lines. Thus, spatial phase alignment, i.e., spatially phase aligned pixels, is an inherent characteristic of the NTSC system. This allows the operation described above in terms of both line and frame combing. For example, pixel p.sub.1 in field I has opposite polarity chroma as pixel p'.sub.1 in the third field (III). Correspondingly, pixel p.sub.2 on the succeeding line in field I has the opposite phase chroma from that of pixel p.sub.1 in the same field. As shown, pixels lined along a diagonal from lower left to upper right have chrominance of the same polarity (.phi. or .phi.+.pi.). Pixels along a diagonal from the upper left to the lower right alternate in a chrominance polarity. Vertically adjacent pixels in the vertical columns also alternate in polarity as do horizontally adjacent pixels in the horizontal rows.
The digital television system of the preferred embodiment uses a 4 Fc sampling signal locked to the chroma burst frequency, Fc. This is illustrated in FIG. 7. Because of this sampling method, it will be seen that alternate chroma samples are one hundred eighty degrees (180.degree.) out of phase. As shown by example in FIG. 7, sample x and sample y (alternate samples) are opposite in phase.
Generally speaking, a frame comb will provide perfect separation if no temporal change (e.g. motion) occurs in the image. In fact, absent such change or motion, frame (three-dimensional) combing provides the best luma/chroma separation of the three discussed methods. The presence or absence of such change is detected in a two-step process as shown in FIGS. 8 through 12.
Referring now to FIG. 8 therein is shown a diagrammatical representation of selected pixels from three succeeding fields (I, II, III) such as those shown generally in FIG. 5. The diagram of FIG. 8 shows these pixels in a time-space relationship. Adjacent pixels in each display line are shown along an axis from lower left to upper right with each succeeding pixel being up and to the right from previous pixels. Adjacent lines in each field are shown vertically displaced by a time equal to 1/F.sub.H. Interlaced lines in succeeding fields are shown horizontally displaced by a time equal to 1/F.sub.V. The same line in succeeding fields is represented as being displaced horizontally by a time equal to 2/F.sub.V. Adjacent pixels in the same display line are displaced by a time factor of 1/4 F.sub.C.
FIG. 8A illustrates the orientation between the horizontal, vertical and time axis for the pixel diagram of FIG. 8, as described above. The units for each (1/4.sub.Fc, 1/F.sub.H and 1/F.sub.v, respectively) are also shown.
For clarity purposes, each of the pixels shown in FIG. 8 has been assigned an alpha numeric reference that is indicative of its relationship to the current pixel of the current display line. Because as shown in the diagram, time advances to the right and subsequent scan lines fall vertically below previous scan lines and subsequent pixels are displaced to the right of previous pixels in a raster scan, the current pixel in FIG. 8 is the third pixel shown in line 3 of field III. This pixel has been labeled "A". The previous pixel has been labeled "B", and the next previous pixel has been labeled "C". The corresponding pixels in the previous display line of the same field have been labeled "D", "E" and "F", respectively. Similarly, the corresponding pixels in the adjacent, interlaced line of the previous field have been labeled "A'", "B'" and "C'", respectively. Likewise, the corresponding pixels in the previous display frame have been labeled "A"", "B"", "C"", "D"", "E"", and "F"". These labels are used for convenience throughout the following discussion.
Those pixels that are involved in the comb filtering have been shown in cross-hatching. Those adjacent pixels that are used for motion detection are shown without cross-hatching.
In particular, in the first step, motion is detected by comparing pixels in corresponding fields. This step comprises developing four vectors K and L (from field III) and M and N (from field I) and determining whether their resultants occupy a predetermined region of spatial convergence. Referring to FIG. 8, all four vectors are derived by subtracting pixels corresponding to chroma samples of like phase. Thus, vector M is derived by subtracting pixel 3 of line 3 from pixel 1 of line 1, both pixels being in field I and both having the same chroma phase (pixel F"-pixel A"). Likewise, vector N is derived by subtracting pixel 1 in line 3 from pixel 3 in line 1 in field I (D"-C"). This vector will have the opposite phase since both pixels are 180 degrees out of phase with the pixels used to derive vector M. Correspondingly, vector K is derived by subtracting pixel 3 of line 3 from pixel 1 of line 1 in field III, both pixels having the same chroma phase (F-A). Vector L is derived from subtracting pixel 1 in line 3 from pixel III in line 1, also in field III and both having the same chroma phase (D-C). As with vectors M and N, vectors L and K will have the opposite phase since the pixels used to derive vector K are 180.degree. out of phase with the pixels used to derive vector L.
FIG. 9 shows the resulting vector-diagram. Vectors M and L are of opposite polarity and represent corresponding portions of a displayed image one frame apart. Vectors N and K have similar characteristics. Thus, if no spatial motion has occurred during the frame, the resultant of vectors M and L and the resultant of vectors N and K will substantially converge to zero. To account for noise and the like a predetermined region of convergence is established (through the use of vector filters). If the magnitude components of the resultant vectors fall within this region, it is assumed that no spatial motion has occurred. In the case of no spatial motion, and subject to satisfaction of the additional condition described below, the frame comb technique will be used for luma/chroma separation.
In addition to the foregoing, a second step tests for temporal convergence in determining whether frame comb luma/chroma separation should be used. Temporal convergence is detected by deriving vectors P, Q, U and V. Vector P represents the difference between like-phase pixels (e.g., L-C) in corresponding lines of adjacent frames, e.g., line 3 in fields I and III (C"-A). Vector V represents the difference between like-phase pixels in adjacent lines in adjacent fields (e.g., line 2 in field II and line 3 in field III) (A-C'). Vectors Q and U are formed in a like manner for like phase pixels having the opposite chroma polarity (e.g., L+C) from those used to for vectors P and V (C-A") and (A'-C). These vectors therefore represent the temporal difference between fields III and I and fields III and II.
Referring now to FIG. 10, therein is shown the resulting vector diagram. Vectors U and Q are of opposite polarity and represent corresponding portions of displayed image one field apart and one frame apart, respectively. Vectors P and V have similar characteristics. Thus if no spatial motion has occurred through fields I, II, and III, the resultant of vectors V and P and the resultant of vectors U and Q will substantially converge to zero. To account for noise and the like, a predetermined region of convergence is established (through the use of vector filters). If the magnitude components of the resultant vectors fall within this region, the system assumes that no spatial motion has occurred. In the case of no spatial motion, frame combing is permitted.
FIG. 11 illustrates in block diagram form a circuit for deriving the eight vectors referred to above.
FIG. 12 illustrates a circuit for processing the derived vectors to determine whether frame comb luma/chroma separation should be used or another separation technique should be used.
Referring more specifically to FIG. 11, five input signals are provided to the circuit for deriving the eight vectors L, K, M, N, P, Q, U and V. The five input signals represent the following (with specific examples for line 3, field 3 as the current display line):
(1) D is the previous line (line 1) of the current field (field 3);
(2) A is the current line (line 3) of the current field (field 3);
(3) A' is the interlaced line (line 2) of the previous field (field 2);
(4) A" is the line corresponding to the current line (line 3) of the previous frame (field 1); and
(5) D" is the line (line 1) previous to the current line (line 3) of the previous frame (field 1).
The circuit further includes a plurality of two-pixel delay devices Z.sub.-2 and a plurality of summers. The inputs of the summers are marked with the corresponding pixels from FIG. 8 to illustrate how the vectors are derived. More particularly, a first pixel delay circuit receives the D pixel signal as does a first summer 102. The first two-pixel delay circuit 100 provides a signal to second summer 104. Once the signal has been delayed by the two-pixel period, it corresponds to pixel from FIG. 8. At its negative input, the second summer receives the current pixel (A) information from the A signal which corresponds to pixel 108 in FIG. 8. As shown in FIG. 8, the output of the second summer 104 then represents the vector L on an output line 110.
A second two-pixel delay circuit 112 receives the current line information from the A pixel signal. Accordingly, the output of the second two-pixel delay circuit 112 represents the pixel 114 from FIG. 8. The first summer 102 then subtracts this pixel value from received signal corresponding to the value of pixel 116 from FIG. 8. The output of the first summer on a line 118 then represents the vector K.
A third two-pixel delay circuit 120 receives the A' pixel signal and provides, as a positive input, a signal corresponding to pixel 122 from FIG. 8 to a third summer 124. The third summer 124 then subtracts the current pixel value, corresponding to pixel 108, from the pixel 122 value to provide the vector V on a line 126. A fourth summer 128 also receives the pixel A' signal as a negative input corresponding to the value of pixel 124 from FIG. 8. It adds to this negative value the pixel value of pixel 114, the output from the two-pixel delay circuit 112, to generate the vector signal U on an output line 130.
A fifth summer 132 receives the A" pixel signal as a positive input. It also receives the output of the two-pixel delay circuit 112 corresponding to the value for pixel 114 as a negative input. The summation of these two signals on a line 134 provides the vector Q.
A fourth two-pixel delay circuit 136 also receives the A" pixel signal to provide an output signal corresponding to the value for the pixel 138 as shown in FIG. 8. A sixth summer 140 receives this value as a negative input and the current pixel value, pixel 108, as a positive input. The summation of these two signals on a line 142 provides the vector P.
A seventh summer 144 receives the output of the fourth two-pixel delay circuit 136 as a positive input. It receives as a negative input the D" pixel signal corresponding to the value for pixel 146 in FIG. 8. The summation of these two on a line 148 provides the vector N.
A fifth two-pixel delay circuit 150 also receives the D" pixel signal to provide at its output a signal corresponding to the value of pixel 152 in FIG. 8. An eighth summer 154 receives this signal as a negative input. The summer 154 combines the negative value for pixel 152 with a positive value for pixel 156 from FIG. 8. The combination of these signals on a line 158 provides the vector M.
It will also be understood that the vectors are derived continuously and are continuously tested for convergence as shown in FIGS. 9 and 10.
Referring to FIG. 12, the eight vectors derived in FIG. 11 are applied to four summers, each of which is coupled to a coring circuit. A typical coring circuit transfer characteristic f(x) is shown in more detail in FIG. 13. For input signals having an absolute magnitude less than a predetermined value x.sub.0, the coring circuit output is 0. For magnitudes exceeding x.sub.0, the coring circuit provides a positive or negative output determined by the symmetrical linear portions of the transfer characteristic. It will be appreciated that the use of coring circuits such as those shown by the characteristics of FIG. 13 implements the desired convergence region described above for the resultant vectors. Choice of the x.sub.0 value for the coring circuits depends upon the noise parameters and desired display characteristics.
The present system can accordingly be adapted to provide improved display picture quality through diminished noise susceptibility. The selection of x.sub.0 governs the noise immunity of the system. Larger values of X.sub.0 may be selected for noisier environments in which deviations in the received video signal indicate noise rather than actual swings in the video content that would otherwise indicate motion in the picture. In less noisy environments, a smaller x.sub.0 value, and hence a smaller convergence zone, may be chosen. Likewise, the choice of the x.sub.0 value determines the level of noise that can be accepted by the system before it must be labeled as having motion with the attendant status changes, such as precluding frame combing for excessively noisy video signals.
The coring circuits that implement the f(x) function shown in FIG. 13 may be any of a multiple number of designs. It is envisioned that this is merely a matter of design choice, as long as the selected circuitry provides the range of x.sub.0 values for accurate operation in expected noise level for the design environment.
Returning to FIG. 12, the L and M vectors are summed at a first summer 160 to provide an input to the first coring circuit 162, and the K and N vectors are summed at a second summer 164 to provide an input to the second coring circuit 166. If no intra-frame motion has occurred, the outputs of both summers will be relatively small values which will be converted to zero value signals by the respective coring circuits. If intra-frame motion has occurred, the outputs of one or both summers will be relatively large and converted to respective non-zero value signals by the coring circuits. The outputs of the two coring circuits are then compared by a comparator 168.
If no intra-frame motion has occurred, both inputs to the comparator are "0" and a "1" output is produced on an output line 170. If motion has occurred, a "0" output is produced. A NAND function may be used to implement this comparison.
Vectors Q and V and vectors P and U are processed in a similar manner through summers 172 and 176 and coring circuits 174 and 178 and a comparator 180. The output of the second comparator is again "1" on a line 182, if no inter-frame motion has occurred and "0" if there has been motion. Frame combing is considered appropriate if the outputs of both comparators is "1" and the signal is in standard for as described below.
Vectors K and L are also coupled to respective chroma band pass filters 184 and 186. The band pass filters 184 and 186 in the preferred embodiment desirably have a bandwidth of f.sub.c .+-.0.5 mhz. Additional coring circuits 188 and 190 normalize the outputs of the band pass filters 184 and 186 as discussed above. This circuitry determines whether line combing is appropriate.
In particular, whenever there is a vertical transition in the chroma component, the use of a line comb will result in chroma impairment in the form of hanging dots. A third comparator 192 tests for this condition by testing the chroma components of vectors K and L which are derived from pixels on successive lines of the current field. Thus, if the magnitude of the chroma components of vectors K and L are both less than a predetermined value established by the coring circuits 188 and 190, the comparator 192 will provide a "1" output on a line 193 indicating that line combing is appropriate. The comparator 192 will also provide a "1" output indicating that line combing is appropriate if both chroma components exceed the predetermined coring level and are equal and the signal is in standard form as discussed below. A "0" output will be provided by the comparator 192 representing that line combing is not appropriate if the outputs of the two coring circuits are different.
Summers 191 and 177 in conjunction with absolute magnitude circuits 195 and 197 and summer 199 provide a similar function with respect to the luma signals. A coring circuit 201 processes this output for use in other video processing circuitry and impresses a normalization range on luma transitions.
Before a final decision as to which luma/chroma separation technique to use is made, one last test is performed. This test determines whether the signal is standard or non-standard. In the preferred embodiment, cooperation with a digital television system is envisioned. A digital system typically outputs a signal which indicates whether a received signal is in standard or non-standard form. A standard signal satisfies the relationship Fc=455 F.sub.h /2, while a non-standard one does not. In the preferred embodiment this signal is entitled "SKEW" and is received on a line 194 by a standard/non-standard decoder 196. The decoder 196 analyzes this property and provides a "1" output on a line 198 if the signal is standard and a "0" output if it is non-standard. In the case of a non-standard signal, the other logic is overridden and one-dimensional filtering techniques are used for luma/chroma separation.
A combinational logic circuit 200 receives the outputs of the comparators 168, 180 and 192 and the decoder 196 on lines 170, 182, 193 and 198, respectively. This logic circuit 200 provides output signals at line 202 indicative of the luma/chroma separation technique to be used. The combinational logic circuit 200 illustrated in FIG. 12 is shown in more detail in FIG. 14.
Referring specifically to circuit of FIG. 14, a first AND-gate 204 receives the output of the comparator 168 on a line 170, the output of the second comparator 180 on a line 182 and the output of the decoder 196 on a line 198. The output of the first AND-gate 204 on a line 206 gives an identification of inter-field and intra-field motion, which if not present indicates the desirability of frame comb separation (i.e., three-dimensional) for luma/chroma separation. This output indicates the desirability of using the field difference and vertical detail signals. A second AND-gate 208 receives the output of the third comparator 192 on a line 193, the output of the decoder 196 on the line 198 and the inverted output of the first AND-gate 204. The second AND-gate 208 provides an output on a line 211 that identifies intra-field motion, which if not present is indicative of the desirability of line comb (i.e. two-dimensional) separation for luma/chroma processing. This output suggests the use of line difference and vertical detail signals. A third AND-gate 212 receives the inverted output of the first AND-gate 204 and the inverted output of the second AND-gate 208 to provide an output signal on a line 214. When the signal on the line 214 is high, it indicates the desirability of one-dimensional separation for luma/chroma processing in a straight filtering technique.
This circuit implements the following logic in determining the priority of luma/chroma separation techniques:
1) If the signal on lines 170 and 182 are "1" (no spatio-temporal motion) and the signal is standard, frame comb is used.
2) If the signal on line 170 or 182 is "0" (motion), but the signal on line 193 is "1" (no vertical chroma transition) and signal is standard, line comb is used.
3) If the signal on line 170 or 182 is "0" (motion), and this signal on line 193 is "0" (vertical chroma transition) filters are used.
4) In any case, if signal is non-standard, filters are used.
It should be understood that the foregoing system is adaptive in an intra-field sense. That is, because luma/chroma separation determinations are continuously made on a pixel-by-pixel basis, different techniques can be employed at different times in a single field. The system will therefore continuously change the luma/chroma separation technique in use depending upon the current pixel information to select that technique most appropriate at any given point in time.
Also, it is envisioned that the general principles of motion detection to control signal processing may be used in systems other than that of the preferred embodiment. For example, a simpler system having the capability of only line combing or filtering could benefit from the same or slightly modified motion detect scheme.
Referring now to FIG. 15, therein is shown a general representation of a digital system for use with the preferred embodiment of the present invention. The video switch 210 may receive the form inputs shown. That is, the envisioned system illustrated in FIG. 15 responds to composite video (CV #1 or CV #2) or to separate chroma (C) and luma (Y) signals. The separate chroma and luma signals may be received in special case applications, such as Super VHS, which supply video signals in this format. The video switch 210 then supplies a corresponding signal to the adaptive comb circuit 214 and a composite video (CV) memory bank 212. The system generally comprises a video switch and analog to digital adaptive luma/chroma separation unit 214.
The memory bank 212 and the adaptive separation unit 214 communicate with a motion detect and memory control 216 in accordance with the present invention as described in more detail above. The output of the motion detect memory control unit 216 controls the adaptive separation unit 214, the output of which is provided to a VPU 218, a DPU 220 and a TPU 222. The output of the adaptive separation unit 214 is also communicated to display circuitry 224. Video signals from the VPU 218 are supplied to a chroma/color difference processing circuit 226. A luma signal from the unit 226 is provided to a luma processing circuit 228, which also receives a video detail signal from the adaptive separation unit 214. The luma processing circuit 228 operates in conjunction with a luma memory bank 232. The output of the luma processing circuit 228 is provided to the display circuitry 224. Correspondingly, the chroma (color difference) output of the circuit 226 is provided to the display circuitry 224.
Referring now to FIG. 16, therein is shown an implementation according to a preferred embodiment of present invention for the three-dimensional adaptive luma/chroma separation circuit 214, shown previously in FIG. 15. This separation circuit further generates the appropriate vertical detail signal for each implemented separation techniques and also provides input/output compatibility for special case application, such as Super VHS format equipment and signals.
More specifically, the system receives a digital video signal at a time multiplexer 300. For purposes of this discussion, the current pixel values (B) represents the received signal. The multiplexer 300 provides a signal then to a gray to binary code converter 302. The output of the converter 302 is supplied on a line 304 having multiple taps, each of which ultimately communicates to one of three multiplexors, a luma multiplexer 306, a chroma multiplexer 308 or a vertical detail multiplexer 310. In one rare instance, the signal on the line 304 may be provided directly to an input multiplexer 312. The system of FIG. 16 also receives inputs from a memory such as that of the composite video (CV) memory bank 212 shown in FIG. 15.
Particularly, the system receives a signal on a line 314 which represents the corresponding pixel of the same line one frame previously B". The system also receives on a line 316 a signal representative of the corresponding pixel in the adjacent, interlaced line in the previous field B'. Finally, the system receives on a line 318 signal representative of the corresponding pixel in the previous line E of the same field.
A system of summers and scaling circuits associated with each of the luma/chroma multiplexers provides for the desired frame or line combing or filtering for the selected luma/chroma separation technique. More specifically, a first summer 320 adds the current line signal with the previous line signal to provide an output signal representative of twice luma with chroma being cancelled. This signal is summed with the signal on line 322 representative of the low frequency components of the vertical detail signal at a second summer 324. This adds back in the vertical detail which may have been removed with the chroma components at the first summer 320. The output of the second summer 324 scaled by a circuit 326 to provide input I1 to the luma multiplexer 306. This input represents line combed luma corrected with vertical detail. Another summer 328 receives at one positive input the current line signal. The summer 328 receives at its other positive input the corresponding line signal from the previous frame from the input line 314. This effectively cancels the chroma component and provides a signal having twice luma magnitude to a one-half scaling circuit 330. The output of the scaling circuit 330 provides input I2 to the luma multiplexer 306, thereby representing frame combed luma.
A low pass filter 332 receives the current pixel signal from the input line 304 and provides a low pass filtering. The output of low pass filter 332 forms input I3 to the luma multiplexer 306, thereby representing filtered luma with chroma desirably filtered out. Finally, input I4 to the luma multiplexer 306 represents the current pixel information from the input line 304.
Correspondingly, a summer 334 subtracts the previous display line information on input line 318 from the current line information on input line 304. This cancels luma and provides a signal having twice the chroma magnitude. A scaling circuit 336 scales this signal to provide input I1 to the chroma multiplexer 308. Input I1 thereby represents line combed chroma. Similarly, a summer 338 subtracts the display line information from the previous frame from the current line information on input line 304. This cancels luma and provides a signal at twice normal chroma magnitude. A one-half scaling circuit 340 provides input I2 to the chroma multiplexer 308. Input I2 then represents frame combed chroma.
Input I3 to the chroma multiplexer 308 simply represents the band passed information from line 304 as filtered by a bank pass filter 341. The band pass filter 341 desirably has a pass band of F.sub.c .+-.0.5 mHz as shown. Input I4 to the chroma multiplexer 308 provides a special case situation (Super VHS) from a second gray to binary code converter 342 which receives a pure chroma signal from the time multiplexer 308 on a line 344. Input I5 to the chroma multiplexer 308 represents the line signal from the interlaced line of the previous field supplied on the input line 316. The output of the chroma multiplexer 308 is provided to a band pass filter 346. Desirably, the band pass filter 346 has a band width centered at Fc having a range of plus or minus 0.5 MHz (.+-.0.5 MHz).
Yet another summer 348 generates a vertical detail signal as input I2 to the vertical detail multiplexer 310. The summer 348 subtracts the previous, interlaced line signal on the input line 316 from the current line signal on input line 304. The vertical detail multiplexer 310 also receives a signal corresponding to twice chroma as the output from the line combing summer 334 as its input I1. Finally, the vertical detail multiplexer 310 receives video corresponding to the previous frame, same line, from the input line 314 as its input I3.
Operation of the time demultiplexor 300 depends on user identification of the desired mode. The preferred embodiment envisions a menu-driven display which allows the user to select special case applications, such as Super VHS. Selection of the Super VHS mode reconfigures the time demultiplexor or 300 to provide previously separated, time multiplexed luma (Y) and chroma (c) from input line 301 to dual gray to binary code converters 302 and 342. As noted above, the output of these converters provide input I4 to each of the luma/chroma multiplexors 306 and 308.
Other components shown in FIG. 16 also facilitate the desired luma/chroma separation and vertical detail generation. A field multiplexer 350 operates in conjunction with the input multiplexer 312 to provide signals to the memory tops. The input multiplexer 312 outputs either the current pixel data or the band passed, combed chroma. The field multiplexer 350 corresponding outputs the previous-field pixel data or the output of the vertical detail multiplexer 310.
Similarly, a second vertical detail multiplexer 352 outputs either the vertical detail signal from the vertical detail multiplexer 310 or a low pass filtered derivative of that signal. A peaking filter 354 and limiter 356 then act on the output of the second vertical detail multiplexer 352 to provide a vertical detail output signal on a line 358. These last components tend to limit the Kell effect in the displayed picture. That is, because of limits in the original television camera aperture, vertical resolution is correspondingly limited. The peaking filter and limiter keep the vertical detail signal frequency and magnitude within corresponding limits. Absent this processing, undesirable distortion can occur in the displayed picture. Unfortunately, the displayed picture at the television is limited by the quality of the equipment that originally changes a real-life object into first an optical image and then an electromagnetic signal description of that optical image.
A binary to gray code converter 360 receives the luma output from the luma multiplexer 306. A time multiplexer 362 then receives the converted output from the converter 360. A second gray to binary code converter 364 receives the chroma output from the band pass filter 346 at the output of the chroma multiplexer 308. The first time multiplexer 362 also receives the output of the second binary to gray code converter 364 the second time multiplexer, in response to timing signals at the sampling rate (4 Fc) and at twice the sampling rate (8 Fc) then outputs gray code, time-multiplexed, separated luma and chroma on a line 366 for further processing and ultimately display.
A second time multiplexer 368 receives the binary coded luma and chroma directly from the luma multiplexer 306 and the band pass filter 346 at the output of the chroma multiplexer 308. In response to timing signals at the sampling rate (4 Fc) and at twice the sampling rate (8 Fc), the second time multiplexer outputs time-multiplexed, separated luma and chroma on a line 370. This provides, thus, both source and display compatibility with Super VHS format signals without undue processing--which can only degrade the picture quality through unnecessary processing steps. These Super VHS compatible signals can then be directly digital to analog converted without the further processing typically associated with other video signals, such as those present at the output of the first time multiplexer 362 on the line 366.
Once again, the first vertical detail multiplexer 310 receives at its I1 input a signal representing twice line-combed chroma, at its I2 input a signal representing field difference, and at its I3 input a signal representing the previous frame information. The output of the first vertical detail multiplexer 310 forms input I2 to the field multiplexer 350, input I1 to the second vertical detail multiplexer 352 and the only input to low pass filter 372. The output of the low pass filter 372 forms input I2 to the second vertical detail multiplexer.
In operation, the first vertical detail multiplexer 310 selects the appropriate input to pass as output depending upon whether motion has been detected (dynamic condition) or not (static condition). For the dynamic condition, input I1 is (line difference) is selected. For the static condition, input I2 (field difference) is selected.
The second vertical detail multiplexer 352 operated in conjunction with the first to provide, ultimately a vertical detail signal output at line 358. The second vertical detail multiplexer 352 selects the output of the first vertical detail multiplexer 310 or the output of the low pass filter 372 depending upon whether composite video was received or separate luma/chroma (SVHS) were received and on whether a static or dynamic condition exists. If the system is in the composite video mode and a static condition exists, the second vertical detail multiplexer 352 selects the unfiltered output of the field difference signal. However, if the system is in the composite video mode and a dynamic condition exists, the second vertical detail multiplexer selects the filtered output--which removes the chroma to leave only the luma, line difference from the field.
On the other hand, whenever the system is in the Super VHS mode, the second vertical detail multiplexer 352 always selects the unfiltered output of the first vertical detail multiplexer 310; in this mode, there is no chroma in the signal and, hence, no reason to separate the chroma from the vertical detail signal. (Note that feedback line 322 adds luma back to the line filtered luma that forms input I1 to the luma multiplexer 306--this avoids unwanted vertical filtering the displayed picture when vertical transitions occur.)
Referring now to FIG. 17, therein is shown a table representing the selected inputs for each of the various multiplexers shown in FIG. 16. The multiplexers shown in FIG. 16 respond to the control signals generated by the motion detection circuit and techniques described above. The control signals arrive via internal system wiring represented by output line 202 in FIG. 12. The luma multiplexer 306 receives the control signals at an input line 307 to select the appropriate input. The chroma multiplexer 308 receives the control signals at an input line 309 to select the appropriate input. The vertical detail multiplexer 310 receives the control signals at a control line 311 to provide a vertical detail signal corresponding to the selected luma/chroma separation technique.
A control circuit 305 responds to the control signal on a line 202' (corresponding to line 202 in FIG. 12) and to signals representative of the vertical rate (Fv), the sampling rate (4 Fc) and twice the sampling rate (8 Fc). The control circuit 305 then provides the control signals that selectively switch the multiplexers in the preferred embodiment. The inputs to the control circuit 305 provide it with information as to whether there is motion or not (and correspondingly which separation technique to select and timing information to allow continuous processing on a pixel basis.
In addition to the luma, chroma and vertical detail multiplexers 306, 308 and 310, the control circuitry 305 also manipulates additional circuitry to provide the desired outputs in accordance with the tables of FIG. 17. The central circuit 305 also governs the multiplexing operation for the input multiplexer 312, at a line 313, the field multiplexer 350, at a line 351, and a second vertical detail multiplexer 352 at a line 353. The corresponding control signal outputs from the control circuit 305 are shown at lines 307'-313', 351' and 353'.
The configuration of the control circuit 305 is a matter of design choice and convenience. The operating parameters are given by the table of FIG. 17 and by the timing information provided by the 4 Fc, 8 Fc and Fv timing signals.
The table of FIG. 17 comprises a first section for composite video and a second section for a special case application. It is based on the following conditions:
S (standard): 455/2 F.sub.H =F.sub.c.
NS (non-standard) 455/2 F.sub.H does not equal F.sub.c.
Static: refers to the presence of a motionless pixel.
Dynamic: refers to the presence of motion for the pixel in question.
FIG. 17 thereby shows the desired result and implementation as discussed above. For example, in a static condition (i.e., motionless pixel) with a standard signal, both the luma and the chroma multiplexers will choose the frame comb input. In contrast, in a dynamic condition (i.e., motion for the pixel in question), and a standard input signal, both the luma and chroma multiplexers will choose I1, the line comb input. Finally, in response to any non-standard signal, each of the luma and chroma multiplexers will choose I3 as the appropriate input, thereby providing for filtering of the composite video signal to separate luma and chroma.
Also illustrated by FIG. 17 is an additional feature which provides for display and source compatibility with Super VHS format tapes and equipment. As is known, Super VHS provides a wider bandwidth signal to provide greater detail in the displayed picture. Also, Super VHS provides luma and chroma already separated, and accordingly, the television system has no need to provide additional separation. Accordingly, in such circumstances it is desirable to provide the luma and chroma directly for display rather than attempting to perform luma/chroma separation processing which could only distort the already separated signals. Accordingly, the time multiplexer 300 recognizes the super VHS format and provides the luma to the first gray to of binary code converter 302 and provides the chroma to the second gray to binary code converter 342. Thus, as shown in the Y/C section of the table of FIG. 17, input I4 to each of the luma multiplexer 306 and the chroma multiplexer 308 will represent separated luma and chroma, respectively, in the presence of the super VHS format signal. In such a condition, each of the multiplexers will choose the I4 input and thereby pass the already separated luma and chroma for further processing and display.
Although the preferred embodiment has been described in terms of a system for use with the NTSC broadcast standards, it is envisioned that the present invention may be used with equal facility with other broadcast standards. The fundamental operation of the luma/chroma filtering, line combing and frame combing will remain predictable, as long as the relationship between luma and chroma in the broadcast signal remains predictable. Desirably, the chroma phase/polarity relationship that is relied upon in generating the motion-indication variables will remain regardless of the broadcast standards. Modification and alteration of the preferred embodiment to accommodate other broadcast signal standards is deemed to be within the skill of those in the art who have become familiar with the teachings of the application.
Claims
What I claim is:
1. A method for detecting motion in a composite video signal for display on a television receiver or monitor comprising: generating a first set of vectors indicative of picture content of a single frame in a received composite video signal; resolving said first set of vectors; comparing said resolved vectors with a preselected convergence range; and providing an indication of intra-frame motion when said resolved vectors fall outside said preselected convergence range.
2. The method of claim 1 further including providing an indication of no intra-frame motion when said resolved vectors fall within said preselected convergence range.
3. The method of claim 1 further including: generating a second set of vectors indicative of picture content of at least two successive frames in a received composite video signal; resolving said second set of vectors; comparing said resolved second set of vectors with a second preselected convergence range; and providing an indication of inter-frame motion when said second set of resolved vectors fall outside said second preselected convergence range.
4. The method of claim 3 further including providing an indication of no inter-frame motion when said resolved vectors fall within said second preselected convergence range.
5. A method for improving television system response to noise in a transmission environment comprising: generating a first set of signals indicative of variations in picture content within a single frame of a received composite video signal; and combining said first set of signals; generating a first signal indicative of a first method for further processing of said composite video signal when combined signals exceed a preselected value; and generating a second signal indicative of a second method for further processing of said composite video signal when said combined signals do not exceed said preselected value.
6. The method of claim 5 further comprising: generating a second set of signals indicative of variations in picture content between adjacent frames of said received composite video signal; combining said second set of signals; generating said first signal indicative of a first method for further processing of said received composite video signal when at least one of said combined first and second sets of signals exceeds a preselected value; and generating said second signal indicative of said second method for further processing of said composite video signal when both of said combined signals do not exceed said preselected value.
7. A method for use in detecting motion in a television signal including: generating a first signal indicative of spatial variance in a received television signal; generating a second signal indicative of temporal variance in said received television signal; comparing said first signal with a first preselected value; generating a first state signal when said first signal exceeds said first preselected value; comparing said second signal with a second preselected value; and generating said first state signal when said second signal exceeds said second preselected value.
8. The method of claim 7 further including generating a second state signal when said first preselected value exceeds said first signal and said second preselected value exceeds said second signal.
9. A method for detecting motion in a television signal including: generating a first signal indicative of spatial variance in a received television signal, comprising generating at least one pair of spatial vectors representative of variations in content within a frame of a television picture signal; and generating a second signal indicative of temporal variance in said received television signal.
10. The method of claim 9 further including generating a first state signal when either of said first and second signals exceeds a preselected value.
11. The method of claim 10 further including generating a second state signal when neither of said first and second signals exceeds said preselected value.
12. The method of claim 9 further including generating at least a second pair of spatial vectors representative of variations within a second frame of a television picture signal.
13. The method of claim 12 further including: resolving said one pair of spatial vectors representative of variations in content within a frame and said second pair of spatial vectors representative of variations within a second frame to obtain a resolved spatial value; comparing said resolved spatial value with a predetermined value indicative of acceptable system noise; and generating a first output signal when said resolved spatial value is less than said predetermined value.
14. The method of claim 13 further including generating a second output signal when said resolved spatial value is greater than said predetermined value.
15. The method of claim 14 wherein said generating said first signal indicative of temporal variance comprises generating at least one pair of temporal vectors representative of variations in content between two or more fields of a television picture signal.
16. The method of claim 15 further including generating at least a second pair of temporal vectors representative of variations between two or more fields of a television picture signal.
17. The method of claim 16 further including: resolving said one pair of temporal vectors representative of variations in content between two or more fields and said second pair of temporal vectors representative of variations between two or more fields to obtain a resolved temporal value; comparing said resolved temporal value with a predetermined value indicative of acceptable system noise; and generating a first output signal when said resolved temporal value is less than said predetermined value.
18. The method of claim 17 further including generating a second output signal when said resolved temporal value is greater than said predetermined value.
19. The method of claim 18 wherein said television picture signal comprises a composite video signal.
20. The method of claim 18 further including generating a control signal indicative of whether said first output signal or said second output signal has been generated.
21. A method for detecting motion in a television signal including: generating a first signal indicative of spatial variance in a received television signal; and generating a second signal indicative of temporal variance in said received television signal comprising generating at least one pair of temporal vectors representative of variations in content between two or more fields of a television picture signal.
22. The method of claim 21 wherein said television picture signal comprises a composite video signal.
23. The method of claim 21 further including generating at least a second pair of temporal vectors representative of variations between two or more fields of a television picture signal.
24. The method of claim 23 further including: resolving said one pair of temporal vectors representative of variations in content between two or more fields and said second pair of temporal vectors representative of variations between two or more fields to obtain a resolved temporal value; comparing said resolved temporal value with a predetermined value indicative of acceptable system noise; and generating a first output signal when said resolved temporal value is less than said predetermined value.
25. The method of claim 24 further including generating a second output signal when said resolved temporal value is greater than said predetermined value.
26. A system for detecting motion in a television picture signal comprising: means for generating a first signal indicative of spatial variance in a received television signal by generating a set of vectors representative of variation in picture content within one or more frames of a television picture signal; and means for generating a second signal indicative of temporal variance in said received television signal.
27. The system of claim 26 further including means for generating a first output signal when either of said first signal and said second signal is greater than a predetermined value.
28. The system of claim 27 further including means for generating a second output signal when neither of said first signal and said second signal are greater than said predetermined value.
29. The system of claim 26 wherein said television picture signal is a composite video signal.
30. The system of claim 26 wherein said means for generating a set of vectors comprises means for generating at least one pair of spatial vectors representative of picture content within a frame of a received television picture signal.
31. The system of claim 30 further including means for generating a second pair of spatial vectors representative of picture content within a second frame of said received television signal.
32. The system of claim 31 further including means for resolving said one and said second pairs of spatial vectors to generate a resolved spatial value.
33. The system of claim 32 further comprising means for generating a first output signal when said resolved spatial value exceeds a predetermined value.
34. The system of claim 33 further comprising means for generating a second output signal when said resolved spatial value does not exceed said predetermined value.
35. The system of claim 34 wherein said means for generating a second signal indicative of temporal variance comprises means for generating a set of vectors representative of variation in picture content between two or more fields of a television picture signal.
36. The system of claim 35 wherein said means for generating a set of vectors representative of variation in picture content between two or more fields comprises means for generating at least one pair of temporal vectors representative of picture content variation between two fields of a received television picture signal.
37. The system of claim 36 further including means for generating a second pair of temporal vectors representative of picture content variation between two fields of said received television signal.
38. The system of claim 37 further including means for resolving said one and said second pairs of temporal vectors to generate a resolved temporal value.
39. The system of claim 38 further comprising means for generating a first output signal when said resolved temporal value exceeds a predetermined temporal value.
40. The system of claim 39 further comprising means for generating a second output signal when said resolved temporal value does not exceed said predetermined temporal value.
41. A system for detecting motion in a television picture signal comprising: means for generating a first signal indicative of spatial variance in a received television signal; and means for generating a second signal indicative of temporal variance in said received television signal by generating a set of vectors representative of variation in picture content between two or more fields of a television picture signal.
42. The system of claim 41 wherein said television picture signal is a composite video signal.
43. The system of claim 41 wherein said means for generating a set of vectors comprises means for generating at least one pair of temporal vectors representative of picture content variation between two fields of a received television picture signal.
44. The system of claim 43 further including means for generating a second pair of temporal vectors representative of picture content variation between two fields of said received television signal.
45. The system of claim 44 further including means for resolving said one and said second pairs of temporal vectors to generate a resolved temporal value.
46. The system of claim 45 further comprising means for generating a first output signal when said resolved temporal value exceeds a predetermined value.
47. The system of claim 46 further comprising means for generating a second output signal when said resolved temporal value does not exceed said predetermined value.
48. A system for improving noise sensitivity in composite video signal processing comprising: first means for combining signals representative of picture content in a display field to generate one or more signals indicative of intra-field motion; second means for combining signals representative of picture content between two adjacent display fields to generate one or more signals indicative of inter-field motion; and third means for combining signals representative of picture content in two adjacent display frames to generate one or more signals indicative of inter-frame motion.
49. The system of claim 48 further comprising: first resolving means for resolving said one or more signals indicative of intra-field motion; and second resolving means for resolving said one or more signals indicative of inter-field motion and said one or more signals indicative of inter-frame motion.
50. The system of claim 49 wherein said first resolving means further comprises means for generating a first output signal when said resolved signals exceed a preselected value and a second output signal when said resolved signals do not exceed said preselected value.
51. The system of claim 50 wherein said second resolving means further comprises means for generating a first output signal when said resolved signals exceed another preselected value and a second output signal when said resolved signals do not exceed said another preselected value.
52. A system for processing television picture signals based on actual picture content to allow improved noise discrimination comprising: first combining means for combining a vector indicative of the difference between pixels of a like, first chroma polarity in a first display field with a vector indicative of the difference between pixels of a like, but opposite, chroma polarity in another display field to produce an output; second combining means for combining a vector indicative of the difference between pixels having the same chroma polarity, opposite to said first chroma polarity, in said first display field and a vector indicative of the difference between pixels having the same, first chroma polarity in said another display field to produce an output; means for resolving an output of each of said first and second combining means to a first value if said output has a magnitude less than or equal to preselected magnitude; and comparison means coupled to said means for resolving for generating a first output signal when the output of each of said combining means has been resolved to said first value indicative of an acceptable variance in picture content within said first and another display fields and for otherwise generating a second output signal.
53. The system of claim 52 further comprising: third combining means for combining a vector indicative of the difference between pixels having a first, same chroma polarity located in two different display fields and another vector indicative of the difference between pixels, at least one being located in each of said two different display fields, having the same chroma polarity, opposite to said first chroma polarity to produce an output; fourth combining means for combining a vector indicative of the difference between pixels having a chroma polarity opposite to said first polarity located in two different display fields and another vector indicative of the difference between pixels, at least one being located in each of said two different display fields, each having said first chroma polarity to produce an output; means for resolving an output of each of said third and fourth combining means to a first value if said output has a magnitude less than or equal to a preselected magnitude; and comparison means coupled to said means for resolving for generating a first output signal when the output of each of said combining means has been resolved to said first value indicative of an acceptable variance in picture content within said first and another display fields and for otherwise generating a second output signal.
54. The system of claim 53 further comprising logic means associated with said comparison means for generating a first control signal when both of said comparison means generate said first output and a second control signal when either of said comparison means generates said second output signal.
55. The system of claim 53 further comprising: means for separating the chroma component from said vector inputs to said first combining means and from said vector inputs to second combining means; and means for comparing said separated chroma components from said vector inputs to generate a first output signal when said separated chroma components are equal in magnitude and a second output signal when said separated chroma components are not equal in magnitude.
56. The system of claim 55 wherein said means for comparing comprises: means for normalizing said separated chroma components to a first value whenever they have a magnitude within a preselected range of magnitudes; and means responsive to said means for normalizing to generate a first output signal when each of said separated chroma components has been normalized and when said separated chroma components are equal in value and to generate a second output signal when either of said chroma components has not been normalized and said separated chroma components are not equal in magnitude.
57. The system of claim 56 further comprising logic means associated with said comparison means and said means responsive to said means for normalizing for generating a first control signal when both of said comparison means and said means responsive to said means for normalizing generate said first output and a second control signal when either of said comparison means or said means responsive generate said second output signal.
58. A system for generating signals indicative of content in a received television signal comprising: delay means responsive to a plurality of signals representative of one or more pixels in a current display line and one or more pixels from other display lines from the current and previous display fields; and combining means responsive to said plurality of signals and to said delay means for providing a plurality of combined output signals.
59. The system of claim 58 wherein said delay means comprises: a first delay circuit for receiving a signal indicative of a first pixel from a current display line in a current display field; a second delay circuit for receiving a signal indicative of a second, corresponding pixel from a previous display line in said current display field; a third delay circuit for receiving a signal indicative of a pixel corresponding to said first pixel from a display line in a previous display field; a fourth delay circuit for receiving a signal indicative of a pixel corresponding to said first pixel from a display line in a previous display frame; and a fifth delay circuit for receiving a signal indicative of a pixel corresponding to said second pixel from a display line in a previous display frame.
60. The system of claim 59 wherein each of said delay circuits comprises a two pixel delay circuit.
61. The system of claim 59 further wherein said combining means comprises: first summer means adapted to receive said signal indicative of said first pixel in said current display line at a first input and the output of said second delay circuit at a second input to generate an output signal representative of the difference between tow pixels in said current display field, said two pixels having a first, like chroma polarity; second summer means adapted to receive said signal indicative of said first pixel in said current display line at a first input and the output of said third delay circuit at a second input to generate an output signal representative of the difference between two pixels in said separate display fields, said two pixels having a first, like chroma polarity; third summer means adapted to receive said signal indicative of said second, corresponding pixel from another display line in said current display field at a first input and the output of said first delay circuit at a second input to generate an output signal representative of the difference between two pixels in said current display field, said two pixels having a like chroma polarity, opposite to said first polarity; fourth summer means adapted to receive said signal indicative of said pixel corresponding to first pixel from a previous display field at a first input and the output of said first delay circuit at a second input to generate an output signal representative of the difference between two pixels in said separate display fields, said two pixels having a like chroma polarity, opposite to said first polarity; fifth summer means adapted to receive said signal indicative of a pixel corresponding to said first pixel from a display line of a previous display frame at a first input and the output of said first delay circuit at a second input to generate an output signal representative of the difference between two pixels in said separate display frames said two pixels having a like chroma polarity, opposite to said first polarity; sixth summer means adapted to receive said signal indicative of said first pixel in said current display line at a first input and the output of said fourth delay circuit at a second input to generate an output signal representative of the difference between two pixels in said separate display frames, said two pixels having a first, like chroma polarity; seventh summer means adapted to receive said signal indicative of a pixel corresponding to said second pixel from a current display line of a previous display frame at a first input and the output of said fourth delay circuit at a second input to generate an output signal representative of the difference between two pixels in another display field, said two pixels having a first, like chroma polarity; and eighth summer means adapted to receive said signal indicative of a pixel corresponding to said first pixel from a display line of a previous display frame at a first input and the output of said fifth delay circuit at a second input to generate an output signal representative of the difference between two pixels in said separate display fields, said two pixels having a like chroma polarity, opposite to said first polarity.
62. The system of claim 58 wherein said generated signals indicative of content in a received television signal are used for improving noise immunity in a displayed picture.
63. The system of claim 58 wherein said generated signals indicative of content in a received television signal are used for detecting motion in a displayed picture.
Patent Citations (10)
| Patent | Date | Inventor | Cited By |
|---|---|---|---|
| US4639767 | 1987-01-01 | Suzuki | |
| US4641180 | 1987-02-01 | Richter | |
| US4651206 | 1987-03-01 | Ohki | |
| US4739390 | 1988-04-01 | Achiha et al. | |
| US4745458 | 1988-05-01 | Hirano et al. | |
| US4754322 | 1988-06-01 | Okuda et al. | |
| US4853775 | 1989-08-01 | Rouoaris | |
| US4868650 | 1989-09-01 | Weckenbrock | |
| US4891699 | 1990-01-01 | Hamada et al. | |
| US4901145 | 1990-02-01 | Harradine et al. |