Video Image Processing Systems

This invention relates to improvements in video image processing systems particularly those which include the combination of two or more incoming video signals under the control of a control signal.

Three prior art systems for this type of processing are shown in FIGS. 1-3. Although the processing is carried out in a similar manner by all three the inputs and outputs of each system are different.

In video graphics equipment there are two requirements that can be met by this type of processor, the need to produce new images under operator control to give the effect of painting and to combine signals from two or more sources of image signals so that the output image can contain areas of image signals from one or all the sources. The circuit shown in FIG. 1 is a so-called brush processor, described in Ser. No. 326,293 (which is incorporated herein by reference), used in Quantel's video graphic system to create new images. In this system the image is produced by combining new video signals representing luminance or chrominance with signals stored in a framestore in proportions determined by a control signal K. The luminance and chrominance signals to be used are chosen by the operator as is also the notional artists implement to be used in the creation. Command signals can be input to the system by the operator by the use of a stylus and touch tablet to identify image points at which `paint` is to be `applied`. The control signal K for each identified image point is produced at the output of multiplier 2 in FIG. 1 and results from the multiplication of first a brush signal, which is a signal relating to the distribution power of the notional implement chosen by the operator, second a signal relating to the pressure applied by the operator to the stylus and third a stencil signal. The operation of a stencil in this system is analogous to that in the conventional artists equipment. Where the operator has not chosen to use a stencil this signal will be set to 1 and so K will be simply brush signal times pressure. K will always be a value between 0 and 1. The processing is done picture point by picture point but K may be precalculated for each picture point as described in co-pending U.S. application Ser. No. 851,110, however this will not effect the operation of this circuit.

Referring only to luminance signals, although similar processing is applied to chrominance signals, the output is produced by multiplying the incoming luminance chosen by the operator by K in multiplier 3 and adding this to the luminance times 1-K in 5. The luminance applied to 5 is the luminance generated for the particular picture point from previous operations of the system. It will be obvious that the output is then K L in +(1-K) L store, where L in is the incoming luminance and L store is the luminance stored at that point. It is found that this processing gives a very realistic image.

FIG. 2 shows a processor which can combine two picture sources in a way which produces an output image which contains different areas of each image as described in U.S. application Ser. No. 457,098 (which is incorporated herein by reference). Where one image is moved relative to the second, parts of the that image may be made to appear as they move in front of objects in the stationary image. Each picture source is provided with a stencil signal which consists of signals for each picture point having a value between 0 and 1 and these signals are multiplied together to provide a control signal k at the output of multiplier 7. The control signal k is again a digital signal with a value between 0 and 1 and is used to determine the proportions in which the two picture signals are combined. It is to be understood that in this system also separate processing paths will be provided for luminance and chrominance signals. In this system, means are provided for displacing one set of the picture signals, and the corresponding stencil signals, relative to the other set, and for allowing part of one image to be moved around a second image by the operator until the correct position is found when the composite image can then be processed to be part of the final image.

The output from adder 11 in FIG. 2 is kP1+(1-k) P2 where P1 is a signal from the first picture source and P2 is a signal from the second picture source. The circuit components needed to achieve this are the same components as in 2 to 6 FIG. 1 arrangement but the circuits are not usually combined. By calculating the stencil signals for each picture source it can be provided that where those areas in picture 1 are to appear on the output image k will be large and vice versa, where picture 2 is to appear. Additional circuitry is of course provided to produce the addressing and to move one picture relative to the second.

The prior art system shown in FIG. 3 is a processor used in still store systems which may be part of a video graphics system. In these systems one facility that is desirable is to be able to view the contents of one store and at the same time preview the contents of a second store and to be able to cut or fade from one to the other. The control value K in the processing in this case can be used then to achieve fade by gradually changing from 1 to 0 or cut by changing more abruptly. Two still pictures are stored in framestores 12 and 13 and then output to the multipliers 14-17. The factor K is applied as a second input to multipliers 15 and 16 and 1-K to multiplier 14 and 17 and it will be obvious from FIG. 3 that the main output from 20 consists of KP2 +(1-K) P1, and the preview output from 19 consists of KP1, +(1-K) P2. As K changes from 0 to 1 then the image stored in framestore 12 will disappear at the main output from 20 and reappear at the preview output from 19. All three circuits described operate on a point by point basis and with suitable additional components can provide a displayed image.

Although the components used in these prior art systems are well known, the implementation of these processors in hardware can be very costly. The processor for the still store is rather more expensive than the other two because of the large number of multipliers involved.

The aim of the present invention is to provide a means for carrying out these three types of processing which is less expensive in terms of components.

According to the invention there is provided, a video processing circuit comprising:

(a) at least two sources of video signals representing different pictures;

(b) a source of signals representing brush shape;

(c) a video graphic means including a stylus device for producing a control signal which is a function of said brush shape signals and pressure on said stylus;

(d) a source of graphic signals for points on a picture designated by said stylus device;

(e) a source of stencil signals;

(f) framestore means having a writing input and a reading output;

(g) a processing circuit having two video inputs for two video signals L.sub.1 and L.sub.2 and a control input for a control signal K and arranged to produce output signals representing K L.sub.1 +(1-K) L.sub.2 ;

(h) selector means for conditioning the system to operate selectively in any one of a plurality of different modes including;

(1) a mode having said source of graphics video signals connected to one video input of said processing circuit, said output of said framestore connected to the second video input of said processing circuit, the output of said processing circuit being written into said framestore,

said control signal of said video graphic means connected to said control input of said processing circuit, and (2) a second mode having said two sources of video signals connected to said two video inputs of said processing circuits, said control input of said processing circuit connected to said source of stencil signals.

Further according to the present invention there is provided a video processing system comprising, input means for video signals from a plurality of sources,

input means for a plurality of control signals, selector means for selecting video signals from more than one source and for selecting one or more control signals, combining means for combining said selected control signals to produce a combined control signal, processing means for combining the selected video signals in proportions determined by said combined control signal, said processing means being capable of operating in more than one mode, one mode producing one video output which is the combination of two or more input signals,

and the second mode producing two video outputs with one output being the input signals combined in the reverse proportions to the combination of the signals in the other output, and means for selecting the mode of operation of the processing means.

One embodiment of the invention will now be described with reference to the following diagrams.

FIG. 1 shows a prior art brush processor.

FIG. 2 shows a prior art combiner.

FIG. 3 shows a prior art processor from a still store system.

FIG. 4 shows one embodiment of the invention.

FIG. 5 shows a second embodiment of the invention.

Reference first to FIG. 5, the processor shown in this figure is capable of achieving all the results produced by the three prior art circuits described above with what can be seen as a considerable saving in components. The input selector 30 selects any of a number of control signals and video signal inputs the selection being dependent on the mode in which the circuit is to operate. The operation of the circuit will now be described for each of the three processes mentioned previously but referring to luminance signals only.

When the circuit is to operate as a brush processor for video graphics systems the operator provides signals via the control 31 which cause input selector 30 to select luminance from the framestore 51 as the luminance 2 (L2) (50) input and new luminance (49) as luminance 1 (L1) input and also selects brush signals which represent graphic parameters of chosen implements, a pressure signal and a stencil signal as the control signals. The operator of the video graphics system causes the image to be generated by `drawing on a touch tablet with a pressure sensitive stylus`. In order to combine the control signals, values of brush (46) and pressure signals (45) for a pixel of the image are input to multiplier 32 before that pixel is present on the input to the processor and the output of multiplier 32 is delayed in delay 33 until the corresponding pixel arrives. The product of brush (46) and pressure (45) is then reapplied to the multiplier 32 with the stencil 1 signal, to form a second product. Luminance values L1 and L2 are presented to the subtractor 34 at this time and the difference L1-L2 is applied to the multiplier 35 to be multiplied by the said second product of brush (46), pressure (45) and stencil 1. It will be understood that where there is no stencil signal and if the pressure signal and the brush are combined before being input to the processor, the brush signal and pressure signals may simply be input at the same time as the luminance signals. The output from multiplier 32 is the control signal K used in the processing and is, say, an 8 bit digital signal representing a number between 0 and 1. The incoming luminance signals are also 8 bit digital signals. Subtractor 34 provides as an output (L1-L2) and this is applied to multiplier 35 as one input while the second input is the K signal. The output of this multiplier is K(L1-L2) and the most significant bits of these output bits pass to adder 38 whilst the remaining pass via switch 41 to component 39 which in this case acts as an adder. Adder 38 receives as a second input the luminance L2 from the store L2 delayed in delay 36, and the output from the adder is KL1+(1-K) L2 which is the new picture point and this can be stored in the framestore 51 at the correct address and can also be viewed on a display 43. The second adder 39 is provided for the reminder because if K is very small information will be lost if only the first 8 bits are used. For the first pass the second input to the adder will be zero and after that will be the output of adder 39. When the value has accumulated beyond the eight bits in this output switch 42 operates to increase the output produced in adder 38 by the correct amount. This is a well known technique for these types of system.

When the circuit of FIG. 5 operates as a brush processor, i.e., in a brush mode, the following labels can be used:

"P" for the input at 45, which is a signal representing pressure which the operator applies to an implement such as a stylus on a tablet;

"B" for the input at 46, which is a signal representing a selected brush shape;

"S" for the input at 47, which is a stencil signal;

"K(brush)" for the output of multiplier 32, which uses the signals labelled P, B and S to derive and output the control signal K, which in this mode (brush mode), is labelled K(brush);

"Ls" for the input at 50, which is the pixel values stored in the framestore and derived therefrom by the processor, which pixel values s are supplied to the negative input of subtractor 34;

"Ln" for the input at 49, which is new pixel values to be supplies to the positive input of subtractor 34; and

"Lc" for the output of adder 38, which is combined pixel values derived by combining the pixel values Ls and Ln in multiplier 35 and adder 38 in accordance with the control signal K(brush).

It will be obvious that the operation of the circuit is much the same as described above when two images are to be combined. In this case the inputs are luminance signals representing two images P.sub.1 and P.sub.2 (at 49 and 50, respectively) and the control signals are the stencil signals (47, 48) corresponding to those images. The images to be combined may in fact have been produced by a video graphics system in the way described above and the stencils could also have been produced in a graphics system. The two stencil signals are multiplied in multiplier 32 to give the control signal K which can be selected by the output selector 40 to form the stencil signal for the resultant image as well as being applied to multiplier 35. The output image (in which a pixel has the value KP.sub.1 +(1-K)P2 may be viewed on a display 43. The facility for using the remainder of the output of 35 is not required in this processing.

When the circuit of FIG. 5 operates to combine two images, i.e., in a picture blend mode, the following labels can be used:

"P1" for the input at 49, to represent the pixel values of a first picture;

"P2" for the input at 50, to represent the pixel values of a second picture;

"S1" for the input at 47, which is a first stencil signal;

"S2" for input at 48, which is a second stencil signal;

"K(blend)" for the output of multiplier 32, which uses the signals labelled S1 and S2 to derive and output the control signal K, which in this mode (picture blend mode), is labelled K(blend); and

"Pn" for the output of adder 38, which is combined pixel values derived by combining the pixel values P1 and P2 in multiplier 35 and adder 38 in accordance with the control signal K(blend).

The operation of the apparatus in FIG. 5 when switched to third mode corresponds to that in FIG. 4. Considering FIG. 4, in order to provide the fade or cut from one image to the other in the output, two stencil signals are multiplied together in multiplier 32 to give the control signal K. These stencil signals are calculated to give the desired effect. Video signals representing two images are input from two framestores to subtractor 34 the output of which is applied to multiplier 35. The output of multiplier 35, which is K(P1-P2), is provided as an input to adder 38 which also receives the signals from the second picture source giving an output of KP1+(1-K)P2. Subtractor 39 receives an input of K(P1-P2) but subtracts this from P1 to give the preview output of KP2 +(1-K)P1. It can be seen that this circuit gives the same outputs as the circuit in FIG. 3 but using fewer multipliers which gives a saving in cost.

When the circuit of FIG. 5 operates to provide the fade or cut from one image to another, i.e., in a cut-and-fade mode, the following labels can be used:

"P1" for the input at 49, to represent the pixel values of a first picture;

"P2" for the input at 50, to represent the pixel values of a second picture;

"S1" for the input at 47, which is a first stencil signal;

"S2" for the input at 48, which is a second stencil signal;

"K(cut)" for the output of multiplier 32, which uses the signals labelled S1 and S2 to derive and output the control signal K, which in this mode (cut-and-fade mode), is labelled K(cut). The same control signal (cut) can be labelled a "cut-and-fade control signal";

"Pm" for the output of adder 38, which is combined pixel values for a main picture Pm derived by combining the pixel values P1 and P2 in multiplier 35 in accordance with the control signal K(cut) and adding the multiplication result to the pixel values P2, as delayed by delay 36, in adder 38; and

"Pp" for the output of subtractor 39, which is combined pixel values for a preview picture Pp derived by combining the pixel values P1 and P2 in multiplier 35 in accordance with the control signal K(cut) and subtracting the multiplication result from the pixel values P1, as delayed by delay 37, in subtractor 39.

To achieve this mode of operation using the FIG. 5 arrangement signals from the control 31 input selector 30 select the correct stencil signals and signals from two luminance stores. Switch 41 is switched to its second position thus connecting the adder/subtractor 39 to the top 8 bit output of multiplier 35 and switch 42 is in its open position to disconnect 38 from 39. In this mode component 39 is forced to act as a subtractor and receives as a second input luminance information from the first picture source. It will be obvious that this configuration is equivalent to that in FIG. 4. The output selector provides the main output on the luminance 1 path and the preview output on the luminance 2 path and these can be delayed separately, at 43 and 52 respectively.

Although described only in terms of luminance signals this system may be made to handle chrominance signals by simply replacing the luminance inputs by chrominance signals.

It is to be understood that by providing different inputs and output signals the processing system of the invention is not limited to the application discussed above.