FI69381C - SIGNALBEHANDLINGSANORDNING - Google Patents

Description

ΓΒ1 f11 ^ ANNOUNCEMENT 69381 lBJ (11) UTLÄGGNINGSSKRIFT 7001 ^ (45) P-tcr.tti iV.y ^ nr.ctty

Pc tent re I,! -: · In t 10 Cl .10CG

(51) Kv.lk. «/ Lnt.CI. 'H Ok N 9/77 (21) Patent application - Pateruansökning 801 ^ 25 (22) Application date - Ansökningsdag 02.05 .80 <H> (23) Starting date - Giltighetsdag 02.05-80 (41) Became Public - Blivit offentlig 12.11 .80

National Board of Patents and Registration Date of display and publication—

Patent and registration authorities '' Ansökan utiagd och uti.skriften pubticerad 3O.O9.85 (32) (33) (31) Claim claimed - Begärd priority j J .05.79 USA (US) 038283 (71) RCA Corporation, 30 Rockefeller Plaza , New York, NY 10022, USA (72) Dalton Harold Pritchard, Princeton, New Jersey, USA (7¾) Oy Kolster Ab (5 ^) Signal Processing Device - Signalbehandlingsanordning This invention relates generally to a color television signal processing device, and more particularly to a novel device which provides an asymmetrical accent of the vertical details of the reproduced image, bringing with it the impression of subjectively pleasing illumination coming from above.

U.S. Pat. No. 4,096,516 (Pritchard) describes a comb coupled device (CCD) comb filter arrangement which is particularly well suited for separating the luminance and chrominance signal components of a combined color image signal. The device has parallel signal paths formed by a long and a short CCD delay line, which have a common clock controller and which end in a common charge adder. In an illustrative embodiment suitable for processing standard NTSC TV signals, the clock frequency is three times the frequency of the color subcarrier (i.e., 3x3.579545 MHz or 10.73835 MHz) and is used to control 2 69381 charge transitions in a long 683 1/2 degree two-phase delay line, and in a short one-degree two-phase delay line. The sum of the charges transmitted through the long delay line and the second short delay line is used to implement a first comb filter, which filter is characterized by band-stop peaks repeated at regular intervals as a half-odd total of half the line frequency; the luminance signal thus processed is obtained from the output of the first charge adder. The sum of the charges transmitted through the long delay line and the second short delay line (which results in phase inversion) is used to implement a second comb filter, which is characterized by band-stop peaks repeated at regular intervals as half-even multiples of the line frequency. After passing a suitable bandpass filter, the chrominance signal thus processed is obtained from the output of the second charge adder. The vertical signal can also be obtained via a suitable bandpass filter from the output of the second charge adder to combine it with the above-mentioned luminance signal, which signal is obtained from the first charge adder.

When the vertical information is thus separately available in a receiver which makes use of the above-mentioned comb filter arrangement, this makes it possible to introduce "vertical enhancement" in the receiver itself, i.e. increasing the vertical content of the luminance component by increasing the apparent vertical sharpness or resolution. The so-called vertical point distortion correctors are well known for transmission operation, in which case such vertical emphasis is used before the signal is transmitted. However, a conventional vertical dot distortion equalizer utilizes a 2-H type comb filter (using two 1-H delay lines) to provide symmetrical vertical enhancement so that both "undercuts" and "overshoots" will be associated with both black and white and white-black vertical scan shifts.

In accordance with the principles of the present invention, the vertical finishing system is implemented so that the vertical finishing information obtained from the 1-H type comb filter is proportionally increased and then combined with the 3 69381 luminance signal obtained from the 1-H comb filter in which the polarity ratio is selected so that produces an asymmetrical vertical highlight where (a) "overshoots" are associated only with black-and-white and (b) "underscores" only with white-black vertical scan transitions. As a result, an apparent improvement in vertical sharpness is achieved by giving the view the impression of illumination from above, with the brightest points at the top of the bright points and shadows below them. The impression is subjectively pleasing and is in line with the viewer’s actual experience that the lighting of the views, whether indoors or outdoors, takes place from above.

The only drawing shows a block diagram of a part of a color TV receiver, in which a vertical highlighting apparatus according to an embodiment of the present invention is shown as a whole.

In the illustrated apparatus, the combined color image signals generated in the video detector of the color television receiver are input to the above.

The combined video input terminal CV of conventional CCD comb filter connections 11 such as those disclosed in the Pritchard patent. In the drawing, the dashed circuit connections can be conveniently made for a single monolithic integrated circuit.

The connections 11 include an attenuator triplet (non-inverting attenuators 13 and 15 and an inverting attenuator 17) that receive the combined video signals from terminal CV for further input to the respective CCD delay lines: short delay line 19 (N degrees); long delay line 21 (N + 682 1/2 degrees); and a short delay line of 23 (N degrees). The delay lines are controlled by a two-phase clock; they are acted upon by the clock signals generated by the clock controller 24 having a signal frequency (10.738635 MHz) three times the frequency of the color subcarrier (3.579545 MHz) which the clock controller receives at the input terminal CL of the appropriate frequency clock control signal. In the illustrative implementation of the connections 11, the short delay lines 19 and 23 each comprise one delay stage, while the long delay line 21 comprises 683 1/2 degrees. The delay difference between the outputs of the delay lines 19 and 21 corresponds to the line frequency (63,555 ps) of the combined video signal, as does the delay difference between the outputs of the delay lines 21 and 23.

4. · 69381

The output signal of the short delay line 19 is summed with the output signal of the long delay line 21 in the first charge adder 25. The output signal of the short delay line 23 is summed with the output signal of the long delay line 2, in the second charge adder. Reference may be made herein to U.S. Patent No. 4,217,605 (J. Carnes et al.), Which describes preferred structures for performing the functions of charge summers (25, 27) and delay lines (19, 21, 23) and for supplying signals to snare lines (19, 21, 23).

If the less delayed output signal of the short delay line 10 is considered to represent the "current line" information under processing 1, it is obvious that the more delayed output signal of the long delay line 21 corresponds to the "previous line" information in terms of time delays 1-H. Similarly, by processing the signals (non-inverted) to the above delay lines, the charge adder 25 realizes the actual summation of the "current line" and "previous line" signals. Such aggregation results in the formation of a comb filter that has band-stop peaks in its frequency response during the half-paired multiples of half the line frequency. The baseband portion of the output signal of the first charge summer 25 thus comprises a combed luminance signal released from the intervening chrominance signal components (but also loses vertical information).

On the contrary, as a result of the different processing of the signals coming to the delay lines 21 and 23, the second charge adder 27 takes care of the efficient subtraction of the adjacent line signals. Such a reduction results in the formation of a comb filter that has band-stop peaks in the frequency response at half-even multiples of the line frequency. The baseband portion of the output signal of the second charge adder 27 thus comprises a combed chrominance signal released from the intervening high frequency luminance signal components (but associated with low frequency vertical signal signals).

The output signal of the second charge adder 27 is input to the sampling / holding and amplifier unit 33 to change the signal so as to provide an increase in the level of the combed, combed, baseband image signal components of the output signal of the adder 27 relative to subsequent clock signals. , their harmonics and associated bandwidth components. As can be seen from the drawing, such signal processing is performed in unit 33 by inverting the polarity of the baseband components by inverting 1. which baseband components are then fed from the above unit 33 to the output terminals (C, C).

The output signal of the first charge adder 25 is also fed to a sampling / holding and amplifier unit (31) which operates in the same way as unit 33. However, the signal is supplied via a delay unit (illustratively comprising two clock-controlled delay stages) 29 to compensate for filter delay , which is explained below. As can be seen from the drawing, the signal processing is performed in the unit 31 in a non-inverting manner and is further fed to the output terminal L of the unit; this is in contrast to unit 33, where the signals are inverted.

The signals coming to the output terminal L are fed to a low-pass filter 35. The passband width of the filter 35 is illustratively 0-4MHz and the filter 35 passes a wideband combed luminance signal while preventing the propagation of higher frequency components associated with the clock signals. The output signal of the low-pass filter 35 is fed to an adder 41 for combining it with the vertical information.

A low-pass filter connected to terminal C 'selectively passes vertical information. The cut-off frequency of the filter 37 (e.g., k, 8MHz) is such that it certainly prevents the propagation of both the chrominance component and the higher frequency components associated with the clock signals. The output signal of the filter 37 is fed to a vertical processor 43 which supplies the vertical signal to an adder 41. It should be noted that the passband of the filter 37 is narrower than the passband of the filter 35 with the result that it causes a greater delay. This delay difference of the filter is compensated in the above-mentioned delay unit 29.

In order to achieve the "vertical enhancement" referred to in the present invention, it is not sufficient that the vertical metering signal input to the adder 41 is amplified in the processor 43 so that the lost vertical metering information is merely restored. Rather, a higher gain than the above gain is necessary in the signal processing of the processor 43 in order to provide the emphasis on the vertical gain. As can be seen from the drawing, the processor 43 does not perform such signal processing invertly. In order to estimate the resulting increase in vertical data, it is appropriate to consider at this point the effects of the polarities between the "current line" and the "previous line" in each sample of the adder 41. Considering, for comparison, the effects of the same polarity, positive, "current line" and "previous line" on the combed luminance signal, the reserve adder 27 can be considered to combine the positive "previous line" information with the negative "current line" information; i. the output signal-li of the adder 27 is the "previous line" information minus the "current line" information. In this case, the inverted output signal of the unit 33 is the "current line" information minus the "previous line" information. Thus, taking into account the non-inverting mode of operation of the processor, the added "current line" information minus the added "previous line" information is input to the vertical signal reception of the adder 41.

Let us now consider what happens in adder 41 when the information of the "current line" differs from the information of the "previous line", i. when there are difficulties in moving in the direction of the vertical sweep. The unchanged "current line" information received from the filter is amplified by an added "current line" from processor 43. The unchanged "previous line" information from the filter 35 is removed by the added "previous line" information of opposite polarity from the processor 43, resulting in a negative conversion of the "previous line". The overall effect is a momentary, pronounced oscillation in the direction of "new" information, i.e. in the direction of the "current line" information. Thus, when the transition in the vertical sweep direction occurs from black to white, a "crossing" in the white direction results. When the transition in the vertical sweep direction occurs from white to black, the result is a crossing in the black direction, i. "Below standard".

Il 7 .......... ··. ·· .. * ..

.......... 69381

The luminance output signal of the adder 41 thus has the previously mentioned asymmetrical vertical enhancement, which, when improving the vertical resolution or sharpness, provides a subjectively pleasant impression of illumination from above. The output signal of the adder 41 is input to the luminance signal processor 53 to generate the input signal Y of the conventional matrix 55.

The output signal from terminal C of the comb filter circuit 11 is fed through a delay unit 39 and an amplifier 45 to the input of a bandpass filter 47 which selectively passes the combed chrominance signal component while blocking low frequency vertical information and high frequency components associated with clock signals. The output signal of the bandpass filter 47 is fed to a conventional chrominance signal processor 49 to generate color difference signals (R - Y, B - Y, G - Y) for the matrix 55. The red, blue and green color signals (R, G, B) generated by the Y signal together with the corresponding color difference signals of the matrix 55 are input to a conventional TV color image display device. Usually, the Y signal is used to adjust the luminance of the reproduced image, while the color difference signals are used to determine its colors (hues and saturations). The delay of unit 39 is added to the delay of the bandpass filter 47 to provide a delay in the chrominance information propagation path that matches the delay of the luminance signal as it travels to matrix 55.

From the above as a result of the effect of polarities, it should be understood that in order to achieve the desired asymmetry of the vertical emphasis described herein, it is appropriate that the input signal containing vertical detail information to the adder 41 be such that the effect of increased t-highlighted "current line" is similar to the effect of an unhighlighted luminance signal passed through the comb filter. It will be seen that to achieve this result, processing units different from the described example can be combined for inverting and non-inverting processing of signals (e.g., units 31 and 33 are non-inverting and processor 43 is inverting; unit 33 and processor 43 are non-inverting and unit 31 is inverting; and so on.).

In a simple implementation of the present invention, in which the vertical 43 processor has a linear amplitude curve, the processor 43 may be a linear amplifier that amplifies the vertical signal. to the extent required to achieve the desired level of enhancement 1. increase of the input signal containing the vertical information in the adder 41 relative to the effect of the signals received from the filter 35. However, in an improved embodiment of the present invention, the processor 43 preferably has a non-linear amplitude curve, such as that described in U.S. Patent 4,245,237 to W. Lagoni ("Adjustable Non-Linear Video Signal Processing Device.") In such an implementation, the processor 43 amplifies the vertical signal by an amount appropriate to However, for amplitudes below said range, the gain is suitable for restoring the verticals, and for amplitudes above said range, the gain is less than the above gain.For describing suitable connections for use in a processor, reference may be made to said Lagnon patent.

The output terminal R 1 of the chrominance signal processor 49 shows the reference frequency of the color subcarrier frequency (e.g., 3.579545 MHz), illustratively generated by a crystal-controlled oxilator that is phase locked to the color synchronizing burst component of the incoming signals. The frequency truncator 51, which corresponds to the reference signals received from the terminal R ', generates the desired clock frequency for supplying clock control to the input terminal CL. Illustrator 51 may illustratively be of the same preferred type used in U.S. Patent No. 4,224,638 (D. Pritchard et al.) And entitled "Frequency Multiplier."

Il

Claims (5)

A signal processing device for highlighting vertical details in a reproduced television image, comprising a first 1H comb filter (with a line), whose output signals represent only the sun of rescue line video signals and preceding line video signals, which signals of a given polarity are measured by a relatively wide band signal path, given bandwidth and gain to an adder, and a second 1H cam filter whose output signals represent only the difference between rescue line video signals and preceding line video signals, which signals are measured to the adder by a relative narrow band having a bandpass limited to only a low frequency portion of the given passband and having an amplification greater than the given gain for at least portions of the difference representing the output signal, the adder being followed by the remainder of the luminance channel, characterized in that the signals from the signal path (31.35) with relatively narrow bands are measured to the adder (41) in such a direction that the effect of the rescuing line video signals on the difference appears at the adder with a given polarity, and that part of the device comprising signal paths (31.35; 33.37) with a relatively wide and relatively narrow band and the adder (41) is designed to produce an asymmetric peak accentuation of the signal at transitions in the direction of vertical scanning, such that a movement toward white that emphasizes the transition is associated only with a transition in the direction of vertical scanning from black to white and a transition to black which emphasizes the transition are only associated with a transition in the direction of vertical scanning from white to black. Device according to claim 1, wherein the television receiver in question is subjected to receiving incoming composite color video signals at a given line frequency and contains a luminance component and a chrominance component which occupies a portion of the frequency band occupied by the luminance component, characterized in that first comb filter means 13 .. · 69381 (19,21,25) exposes incoming Saxon composite color video signals to a first frequency characteristic which exhibits generally discrete rejection notches occurring at odd multiples of half-line frequency and containing means (25) for summing a first version of incoming color video signals having a second version of incoming composite color video signals, which are delayed a line interval relative to the first version, that the second comb filter means (23, 21, 27) exposes incoming composite color video signals to a second frequency characteristic which exhibits regularly distinct rejection. ck that occurs at even multiples of half-line frequency and contains means (27) for summing the second version of incoming composite color video signals with a third version of incoming composite color video signals, which in time respect substantially corresponds to the first version, but is phase reversed relative to this, that the relatively wide band signal path is coupled between the output of the first-mentioned summing means and an input to the signal adder, and the output from the first comb filter means with a first gain, that the signal path (33,37) with a relatively narrow band is coupled between the output of the the latter said summing means and an input to the adder, and includes a low band filter (37) which limits the signal passage to a frequency band substantially beyond the portion of the band occupied by the chrominance component for reproducing at least a portion of the signal output from the second comb filter means. reinforcement that is greater than the first gain, and that one of the signal paths to exclude the other outside its signal reproduction by grid polarity inversion. Device according to claim 2, characterized in that bandpass filter means (47) which are sensitive to the output of the latter mentioned summing means for selectively deriving the chrominance component from the signal output of the second cam filter means, and a means (49) for utilizing the derivative chrominance component for controlling the color tone and saturation of rendered color images.