Dynamic Range Video Black Level Expander

This application is related to, but in no way dependent upon, copending application Ser. No. 290,783, filed 12/28/88, now U.S. Pat. No. 4,937,671, entitled, DIGITAL VIDEO BLACK EXPANDER, in the name of C. Engel and assigned to Zenith Electronics Corporation.

This invention relates in general to digital television systems and in particular to a black expander for a digital video signal in a digital television system.

In the above-described copending application, a digital black expander expands digital video signals that are below an arbitrary breakpoint. That invention, in its preferred embodiment, initially develops a moving average of samples of horizontal video and then samples the horizontal moving average at a field rate to develop field rate samples. The breakpoint is established by adding a constant to the moving average of field rate samples. On the contrary, the present invention method and apparatus ascertain the dynamic range of the digital video signal for each field of video and uses the dynamic range to determine black signal level expansion. In the absence of a dynamic range for the video signal, no black expansion is permitted even though the video signal is below the expansion tilt point. With the system of the invention, black expansion is not performed except under conditions where doing so will provide a benefit to the viewer. The premise is that if the video signal has no dynamic range, or only limited dynamic range, there is no need for black expansion. Indeed, black expansion under uniform signal conditions may be detrimental to the video display. The high frequencies in the luminance signal are removed prior to processing to eliminate rapid transitions which could create annoying flicker in the displayed video.

It is a principal object of the invention to provide an improved video black level expansion system and method.

Another object of the invention is to provide a novel video black level expansion system and method that operates only when the video display will be enhanced.

These and other objects and advantages of the invention will be apparent upon reading the specification in conjunction with the drawings, in which:

FIG. 1 is a graph indicating the operation of a video black expander;

FIGS. 2A and 2B, when combined, are a partial block diagram of a video processor of the invention; and

FIG. 3 is a graph indicating operation of the video black expander of the invention.

Referring to FIG. 1, a typical video characteristic curve 10 is modified for black expansion by changing its slope at a tilt point 12, as is represented by curve portion 14. Maximum and minimum levels indicate the spread or dynamic range of the video signal for reference purposes. Video signals above a threshold level are not subject to expansion whereas signals below that level (particularly at and below tilt point 12) are expanded in the black direction. This characteristic is representative of that produced by the circuit disclosed in the copending application mentioned above, where the tilt point (or breakpoint) 12 is established by adding a constant to the estimated darkest portion of the video signal. The dynamic range of the video signal plays no part in the expansion criteria.

The black level expander of the present invention provides a similar translation characteristic except that the tilt point is based upon the dynamic range of the input digital video signal. Specifically, the tilt point Xt is calculated as the product of a programmable constant K1 and a factor D that represents the difference between Xmax and Xmin. Therefore, Xt equals (K1) D, where D is equal to Xmin-Xmin. Xmax and Xmin, in accordance with the invention, are integrated pixel values that are derived over successive fields of the video signal.

FIG. 2A and FIG. 2B should be positioned with arrows A, B and C in alignment. A source 18 of luminance (LUMA) signal input (Xin) comprises a series of pixels, each represented by 8 bits. The digitized input signal is supplied through a limiter 20 to a vertical average and horizontal low pass filter (LPF) 22 for removing high frequencies. The frequency range of low pass filter 22 is 0 to 1.2 MHz. The frequency limitation removes any rapid transitions in video from the black level expander processor. A practical implementation includes a three line delay with corresponding pixels from each line being averaged by the vertical LPF and supplied as an output to the horizontal LPF. While not essential, this averaging technique prevents misinterpretation of noise on a single line. The output of LPF 22 is coupled to the A inputs of each of a max comparator 24, a min comparator 26, a max multiplexer 28 and a min multiplexer 32. The output of max comparator 24 is applied to the SEL input of max multiplexer 28, the output of which is connected to the D input of a max latch 30.

It will be recognized that, while only a single latch is indicated for simplicity, there are actually eight latches, one corresponding to each bit of the signal Xin. The Q output of max latch 30 is coupled to the B inputs of max comparator 24 and max multiplexer 28. The output of min comparator 26 is similarly coupled to the SEL input of min multiplexer 32, the output of which is connected to the D input of a min latch 34, of which there are also eight. The Q output of min latch 34 is connected to the B inputs of min comparator 26 and min multiplexer 32. Both latches 30 and 34 are clocked only during active video by a signal that is produced by an AND gate 36, the inputs to which are a pixel clock and a horizontal-vertical (HV) blanking signal. A field rate signal is coupled to the reset terminals of max latch 30 and min latch 34, thus resetting the latches for every field of video.

In operation, max latch 30 is reset at the beginning of each video field to all 0's while min latch 34 is reset to all 1's. The 0 value or level of max latch 30 is compared with the value of the first pixel applied to the A input of max comparator 34, which is presumably greater than 0. Comparator 24 therefore selects the A output of max multiplexer 28 so that the first pixel value is coupled to and stored in max latch 30. This process is repeated for each successive pair of pixels and results in a maximum pixel value being stored in max latch 30. A similar process is performed to derive the minimum pixel value, which is stored in min latch 34. Here again, the value or level of the first pixel applied to the A input of min comparator 26 will normally be less than the reset condition of all 1's for min latch 34. The result of the max and min comparisons is that the brightest video pixel and the darkest video pixel values for each video field are stored in max latch 30 and min latch 34, respectively. The output of max latch 30 is coupled to the input of a max integrator 40 and the output of min latch 34 is coupled to the input of a min integrator 42. The max and min integrators 40 and 42 are clocked at a field rate and receive the max and min values from max latch 30 and min latch 34 each field. Each integrator is arranged to add the new maximum (or minimum) pixel value from its respective latch to the existing level in the integrator and to divide the sum by two.

Assuming the max integrator is initially set to zero, the max pixel value (assumed to be 10) from max latch 30 would be added to zero and divided by two to establish the new pixel value of 5 in max integrator 40. In the next video field, the new max latch pixel value (assumed to be 13) is added to the max integrator pixel value of 5 and divided by two to establish a new pixel value of 9 for max integrator 40. The process is repeated for all pixels in the field. A similar process occurs with respect to the min integrator 42. The provision of the integrators effectively desensitizes the dynamic range calculation and precludes black expander operation in response to short term changes in video dynamic range.

The output of max integrator 40 is applied to the positive input of an adder 44 and the output of min integrator 42 is applied to the negative input of adder 44. Effectively, the value of the min integrator 42 output is subtracted from the value of the max integrator 40 output to produce a dynamic range value (D) at the output of adder 44. A multiplier 46 multiplies the dynamic range value D by a factor K1, which is programmable as indicated by the arrow, labelled Xrange, adjacent to multiplier 46. This produces a value (K1) D. Constant K1 is programmable in 16 steps from 1/16 to 1. The output of min integrator 42 is also coupled to the A input of a comparator 48, the B input of which is supplied with a programmable limit value Xlimit. When the minimum value at the output of min integrator 42 is greater than or equal to the limit value Xlimit, comparator 48 develops an output for disabling the black expansion processing system by causing a switch 50 to assume an open position.

With switch 50 in its closed position as shown, the level (K1) D is supplied to the positive input of an adder 52. The value at the output of min integrator 42 is also supplied to a positive input of adder 52 which produces an output, Xt that is equal to (K1) D+Xmin. This value is applied to a negative input of another adder 54. Adder 54 receives, at its positive input, the original digital video signal Xin (from limiter 20). Thus Xt is subtracted from the original signal Xin and applied to another multiplier 56 having a programmable slope factor K2. Slope factor K2 is programmable in eight steps of 1/4 each between 7/4 and 0 as indicated by the adjacent arrow labelled Xslope. The output of adder 56 is therefore:

This signal is applied to a limiter 58 having a break at tilt point Xt and an output that is characterized as follows:

If Xin is less than Xt, the output is K2(Xin-Xt).

If Xin is equal to or greater than Xt, the output is 0.

The input signal Xin is added to the output of limiter 58 in another adder 60 which provides a processed output luminance signal after further conventional limiting in a limiter 62, characterized as follows:

If Xin is less than Xt, the output is K2(Xin-Xt)+Xin.

If Xin is equal to or greater than Xt, the output is Xin.

It is thus seen that for values of Xin that are greater than or equal to the tilt point Xt, the original response curve or transfer characteristic is followed. However, for values of Xin that are less than the tilt point Xt, the response curve is modified such that Xout=K2(Xin-Xt)+Xin. Since Xt=K1 times D+Xmin, the tilt point is a function of the dynamic range of the video signal.

With reference to FIG. 3, any value of Xin that is greater than or equal to 40 will follow curve 70 and produce an output Xout=Xin. For values of Xin less than 40, the response will follow curves having different slopes depending upon the value of the slope factor K2. For example, if K2=0, Xout will always equal Xin, that is, the response curve will be coincident with curve 70. If K2=1 and Xin=20, then Xout=0 as represented by the curve marked K2=1. If K2=7/4, then Xout would be 0 for all Xins up to approximately 24.

Consequently, with the invention, black level expansion is performed only when there is a dynamic range to the video signal. The degree or amount of dynamic range determines the tilt point in a direct fashion such that the larger the dynamic range, the higher the tilt point. When the dynamic range D is zero, the tilt point Xt=Xmin, and no expansion occurs. Thus expansion of signals toward black is performed in accordance with the degree to which an apparent improvement in contrast will be most noticeable to the viewer. When the improvement would not be noticeable, i.e., where there is no dynamic range to the signal or where there is only a signal of minimal dynamic range, expansion is not performed (or only performed in a minimal amount). Consequently, with the invention, black expansion is performed under signal conditions where it may be of benefit to the viewer and not performed under other signal situations.

It is recognized that numerous changes and modifications may be made to the preferred embodiment of the invention by those skilled in the art without departing from its true spirit and scope. The invention is to be limited only as defined in the claims.