JP5119740B2 - VIDEO SIGNAL PROCESSING DEVICE, VIDEO SIGNAL PROCESSING METHOD, VIDEO SIGNAL PROCESSING METHOD PROGRAM, AND RECORDING MEDIUM CONTAINING VIDEO SIGNAL PROCESSING METHOD PROGRAM - Google Patents

Description

  The present invention relates to a video signal processing device, a video signal processing method, a video signal processing method program, and a recording medium on which a video signal processing method program is recorded, for example, a video obtained by decoding streaming data by MPEG (Moving Picture Experts Group) or the like. It can be applied to signal processing. The present invention detects the periodic temporal fluctuation amount of the high-frequency component contained in the input video signal, and changes the characteristics of the video signal processing so as to suppress the periodic temporal fluctuation based on the detection result. By doing so, it is possible to reduce the flicker deterioration more reliably than in the prior art. Further, the flicker deterioration is quantitatively measured by using the flicker deterioration reduction technique.

  Conventionally, in various video devices such as a monitor device, image quality is improved by various video signal processing such as noise reduction processing, sharpness enhancement processing, and contrast enhancement processing. Regarding contrast enhancement processing among these video signal processing, in recent years, in flat display devices and the like, the input video signal is amplified with nonlinear characteristics to enhance contrast, and the average luminance level in the screen of the input video signal is increased. The non-linear characteristics are adaptively changed based on a luminance histogram or the like, and the image quality is improved by further enhancing the contrast.

  In recent years, some video signals are transmitted after being compressed by, for example, encoding processing such as MPEG, and so-called flicker degradation may occur in the video signals transmitted by such data compression ( Yoshiaki Shikaga, Katsunori Aoki, Eisuke Nakasu “Verification of Bidirectional Prediction Effect in Video Coding: Image Quality Evaluation and Analysis”, Television Society Journal, Vol. 50, No. 3, 1996, pp. 391-398) .

  Here, the flicker degradation is a phenomenon in which the high frequency component of the video signal varies periodically according to the picture type, and is caused by a difference in data compression in the I picture, P picture, and B picture. Flicker degradation becomes conspicuous by video signal processing such as noise reduction processing, sharpness enhancement processing, and contrast enhancement processing. Specifically, in the case of contrast enhancement processing, flicker degradation becomes conspicuous when contrast enhancement processing is performed on a low-contrast dark scene in which high-frequency components tend to be noticeable due to the non-linear characteristics that enhance the gradation of dark areas. .

  As a method for preventing the flicker deterioration, Japanese Patent Application Laid-Open Nos. 9-224250 and 2005-260902 conventionally detect the average luminance level difference between frames by performing gamma correction on the luminance signal, and this level difference. There has been proposed a method of correcting the signal level of the luminance signal so as to suppress the noise.

However, when the video signal is simply processed linearly, flicker deterioration can be reduced to some extent by the methods disclosed in Japanese Patent Laid-Open No. 9-224250 and Japanese Patent Laid-Open No. 2005-260902. There is still an insufficient problem in practical use. In particular, the methods disclosed in these publications have a problem that flicker degradation cannot be reduced when video signals are processed by nonlinear characteristics by dynamically switching characteristics according to input video signals, which is a recent video signal processing. is there.
JP-A-9-224250 JP-A-2005-260902 Yoshiaki Shikaga, Katsunori Aoki, Eisuke Nakasu “Verification of Bidirectional Prediction Effect in Video Coding: Image Quality Evaluation and Analysis”, Television Society Journal, Vol. 50, no. 3, 1996, pp. 391-398

  The present invention has been made in consideration of the above points. A video signal processing apparatus, a video signal processing method, a program for a video signal processing method, and a video signal that can reduce flicker degradation more reliably than in the past. The present invention intends to propose a recording medium on which a processing method program is recorded. In addition, a video signal processing apparatus and a video signal processing method for quantitatively measuring flicker degradation using this flicker degradation reduction method are proposed.

In order to solve the above problems , according to one aspect, a video signal processing unit that processes an input video signal to generate an output video signal, and detects a high-frequency component amount included in the input video signal A high-frequency component measurement unit, and a high-frequency component variation feature amount measurement unit that detects a high-frequency component variation feature amount indicating a periodic temporal variation amount of the high-frequency component amount from the high-frequency component amount, and The video signal processing unit varies the characteristics of the video signal processing so as to suppress periodic fluctuations in the high frequency component of the output video signal based on the high frequency component fluctuation feature amount, and the high frequency component The variation feature amount measurement unit extracts a signal component corresponding to a picture type periodic pattern set in the input video signal from the high frequency component amount, and from the high frequency component amount, By bandpass output processing A high-pass output process that extracts a high-frequency component including a signal component and a signal component generated by the band-pass output process is corrected by a high-frequency component generated by the high-pass output process to generate the high-frequency component variation feature quantity Component variation feature value generation processing, wherein the video signal processing includes noise reduction processing, sharpness enhancement processing, or contrast enhancement processing, and the video signal processing unit is indicated by the high-frequency component variation feature amount There is provided a video signal processing device that increases the noise reduction effect in the noise reduction processing or decreases the amplification factor in the sharpness enhancement processing or the non-linearity in the contrast enhancement processing as the time variation amount increases .

  According to another aspect, a video signal processing step of processing an input video signal to generate an output video signal, a high frequency component measurement step of detecting a high frequency component amount included in the input video signal, A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from a high frequency component amount, and the video signal processing step includes the high frequency component variation feature amount measurement step. A characteristic variable step for varying the characteristics of the video signal processing so as to suppress periodic fluctuations in the high-frequency component of the output video signal based on a regional component fluctuation feature quantity, and the high-frequency component fluctuation feature The amount measuring step includes a bandpass output step for extracting a signal component corresponding to a periodic pattern of a picture type set in the input video signal from the high frequency component amount, and the high frequency component amount. A high-pass output step for extracting a high-frequency component including a signal component by the band-pass output step, and a signal component by the band-pass output step is corrected by a high-frequency component by the high-pass output step, so that the high-frequency component variation feature A high-frequency component variation feature value generation step for generating a quantity, wherein the video signal processing includes noise reduction processing, sharpness enhancement processing, or contrast enhancement processing, and the characteristic variable step includes the high-frequency component variation feature A video signal processing method that increases the noise reduction effect in the noise reduction processing or decreases the amplification factor in the sharpness enhancement processing or the nonlinearity in the contrast enhancement processing as the amount of time variation indicated by a quantity increases. Provided.

  According to another aspect, a video signal processing step of processing an input video signal to generate an output video signal, a high frequency component measurement step of detecting a high frequency component amount included in the input video signal, A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from a high frequency component amount, and the video signal processing step includes the high frequency component variation feature amount measurement step. The video signal processing step includes a characteristic variable step for changing the characteristics of the video signal processing so as to suppress periodic fluctuations in a high frequency component of the output video signal based on a band component fluctuation feature amount. The input video signal is varied according to the input video signal, and the input video signal is amplified by nonlinear characteristics, thereby enhancing the contrast of the input video signal and generating the output video signal; Based on the high-frequency component variation feature value, the nonlinear characteristic is brought close to a linear property to suppress the periodic variation, and the characteristic variable step includes the time indicated by the high-frequency component variation feature value. A video signal processing method is provided in which nonlinearity in the contrast enhancement executed by the video signal processing step is reduced as the variation amount is increased.

  According to another aspect, a video signal processing step of processing an input video signal to generate an output video signal, a high frequency component measurement step of detecting a high frequency component amount included in the input video signal, A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from a high frequency component amount, and the video signal processing step includes the high frequency component variation feature amount measurement step. A characteristic variable step for varying the characteristics of the video signal processing so as to suppress periodic fluctuations in the high-frequency component of the output video signal based on a regional component fluctuation feature quantity, and the high-frequency component fluctuation feature The quantity measuring step converts the signal waveform of the signal component related to the periodic time fluctuation to the time fluctuation component of the high-frequency component quantity based on the relationship between the waveform of the filter coefficient of the bandpass filter required for the extraction of the signal component. A bandpass vector phase correlation detecting step for detecting a bandpass vector phase correlation indicating a magnitude of the signal component related to the periodic time variation, and an energy value of the signal component related to the periodic time variation. A synchronous bandpass component generation step to detect, and a fluctuation cycle component occupancy calculation step for calculating a fluctuation cycle component occupancy indicating a ratio of an energy value of the signal component to the high frequency component based on the bandpass vector phase correlation degree; The fluctuation period stability calculating step for calculating the fluctuation period stability indicating the stability of the fluctuation period component occupancy in the time axis direction, and the energy value obtained in the synchronous bandpass component generation step are used as the fluctuation period component occupancy. And a variation feature generation step for detecting the high-frequency component variation feature amount by correcting with the degree of variation and the variation period stability. The video signal processing includes noise reduction processing, sharpness enhancement processing, or contrast enhancement processing. In the characteristic variable step, the larger the time variation amount indicated by the high-frequency component variation feature amount, the larger the time variation amount in the noise reduction processing. There is provided a video signal processing method for increasing a noise reduction effect or reducing an amplification factor in the sharpness enhancement processing or non-linearity in the contrast enhancement processing.

  According to another aspect, a video signal processing step of processing an input video signal to generate an output video signal, a high frequency component measurement step of detecting a high frequency component amount included in the input video signal, A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from a high frequency component amount, and the video signal processing step includes the high frequency component variation feature amount measurement step. A characteristic variable step for varying the characteristics of the video signal processing so as to suppress periodic fluctuations in the high-frequency component of the output video signal based on a regional component fluctuation feature quantity, and the high-frequency component fluctuation feature The amount measurement step is a variation type signal generation step for determining the increase / decrease in the high frequency component amount and outputting a variation type signal, and a picture type periodic pattern set in the input video signal. Based on the determination result of the coincidence determination step for determining whether or not the temporal variation pattern by the variation type signal matches the periodic temporal variation pattern assumed in the high frequency component amount, and the coincidence determination step And a variation feature generation step for processing the high-frequency component amount and detecting the high-frequency component variation feature amount when a coincidence determination result is obtained in the match determination step, and the video signal The processing includes noise reduction processing, sharpness enhancement processing, or contrast enhancement processing. In the characteristic variable step, the noise reduction effect in the noise reduction processing increases as the time variation amount indicated by the high-frequency component variation feature amount increases. Video signal processing method for increasing the frequency or reducing the amplification factor in the sharpness enhancement processing or the nonlinearity in the contrast enhancement processing It is provided.

  According to another aspect, a video signal processing step of processing an input video signal to generate an output video signal, a high frequency component measurement step of detecting a high frequency component amount included in the input video signal, A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from a high frequency component amount, and the video signal processing step includes the high frequency component variation feature amount measurement step. A characteristic variable step for varying the characteristics of the video signal processing so as to suppress periodic fluctuations in the high-frequency component of the output video signal based on a regional component fluctuation feature quantity, and the high-frequency component fluctuation feature The amount measuring step includes a bandpass output step for extracting a signal component corresponding to a periodic pattern of a picture type set in the input video signal from the high frequency component amount, and the high frequency component amount. A high-pass output step for extracting a high-frequency component including a signal component by the band-pass output step, and a signal component by the band-pass output step is corrected by a high-frequency component by the high-pass output step, so that the high-frequency component variation feature A high-frequency component variation feature value generation step for generating a quantity, wherein the video signal processing includes noise reduction processing, sharpness enhancement processing, or contrast enhancement processing, and the characteristic variable step includes the high-frequency component variation feature The video signal processing method of increasing the noise reduction effect in the noise reduction processing or reducing the amplification factor in the sharpness enhancement processing or the nonlinearity in the contrast enhancement processing as the amount of time variation indicated by the amount increases. A program is provided. A recording medium recording the program is also provided.

According to these configurations , a high frequency component amount included in the input video signal is detected, and a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount is detected from the high frequency component amount. Thus, it is possible to measure the amount of signal components related to flicker degradation included in the input video signal. As a result, if periodic fluctuations in the high frequency component of the output video signal are suppressed based on the high frequency component fluctuation feature quantity, the characteristic is dynamically switched according to the input video signal, and the video signal is obtained by nonlinear characteristics. Even when processing, flicker degradation can be reduced.

Also, to detect the high frequency component amount included in the input video signal, from the high-frequency component amount, by detecting a high frequency component variation characteristic amount indicating a periodic time variation of the high frequency component quantity, It is possible to measure a signal component amount related to flicker degradation included in the input video signal and control to reduce the flicker degradation by this high frequency component variation feature amount. Accordingly, the flicker deterioration can be quantitatively measured using a flicker deterioration reduction method.

  According to the present invention, it is possible to further reduce flicker deterioration more reliably than in the past. Further, flicker deterioration can be measured quantitatively.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.

(1) Configuration of Embodiment FIG. 2 is a block diagram showing a video signal processing apparatus according to Embodiment 1 of the present invention. In this video signal processing apparatus 1, an input video signal S1 is subjected to noise reduction processing, sharpness enhancement processing, and contrast enhancement processing by a video signal processing unit 2, and an output video signal S2 is output. The video signal processing apparatus 1 uses the high-frequency component measuring unit 3 and the high-frequency component variation feature amount measuring unit 4 to periodically change the high-frequency component amount D1 and the high-frequency component amount D1 included in the input video signal S1. A high-frequency component variation feature quantity D2 indicating the quantity is detected. Further, the high-frequency component amount D1 and the high-frequency component variation feature amount D2 control the video signal processing in the video signal processing unit 2, thereby reducing flicker degradation and noise reduction processing and sharpness enhancement processing of the input video signal S1. , Contrast enhancement processing.

  In this embodiment, the video signal processing apparatus 1 is constituted by a processor that processes the input video signal S1 by executing a predetermined program. In this embodiment, the program is installed and provided in advance. It may be provided by being recorded on a recording medium such as an optical disk, a magnetic disk, or a memory card, or may be provided by downloading via a network such as the Internet.

  In the video signal processing device 1, the high frequency component measuring unit 3 separates the high frequency component from the input video signal S1, and detects and outputs the high frequency component amount D1 of the separated high frequency component. Here, the high frequency component may be a high frequency component in the image space direction of the input video signal S1, or may be a high frequency component in the time direction of the input video signal. The detection unit of the high frequency component amount D1 may be either a field unit or a frame unit. In this embodiment, the high frequency component measuring unit 3 detects the high frequency component amount D1 by setting the detection unit of the high frequency component amount D1 in units of frames and separating the high frequency component in the image space direction. As a result, the high frequency component measuring unit 3 separates the high frequency component from the input video signal S1 using, for example, a two-dimensional high-pass filter, and obtains the absolute value sum, square sum, etc. of the high frequency component in units of frames. The component amount D1 is detected. Also, by setting the two-dimensional high-pass filter, one type of high-frequency component having a specific property is extracted. In this embodiment, this one type of high-frequency component having a specific property is randomly selected in the time axis direction. It is set to a random noise component having a property to be generated. When separating the high frequency component in the time axis direction, the high frequency component can be separated by obtaining the inter-field difference or inter-frame difference of the input video signal S1. In addition, texture and brightness information may be applied to this specific property instead of the random noise component. When texture is applied, the high frequency component amount D1 detected according to the dispersion value of the pixel value. It is possible to increase / decrease the output. When brightness information is included, the detected high frequency component amount D1 may be increased or decreased according to the luminance value and output.

  Here, when the input video signal S1 is a video signal transmitted by MPEG, as shown in FIG. 3, the high frequency component amount D1 detected by the high frequency component measuring unit 3 is set to the input video signal S1. It will change according to the picture type. In FIG. 3, the I picture, the P picture, and the B picture are indicated by symbols I, P, and B, respectively. Therefore, in FIG. 3, one GOP is composed of 15 frames, the first picture of the GOP is set as the I picture, the 3n + 1th picture from the start is set as the P picture, and the rest is set as the B picture. However, n is an integer.

  In a video signal transmitted by MPEG, a high frequency component amount generally decreases in a B picture due to a code amount allocation amount at the time of encoding, a high selection rate of interpolation prediction in bidirectional prediction, and the like. Also, in the I picture, since the code allocation amount is large, the high frequency component amount relatively increases, and in the P picture, the high frequency component amount becomes an intermediate value between the I picture and the B picture. Although the high frequency component amount in each picture varies depending on the rate control at the time of encoding, the temporal variation of the high frequency component amount is a cause of flicker degradation, and the picture set in the input video signal S1. It will occur periodically depending on the type.

  The high frequency component fluctuation feature quantity measuring unit 4 detects a periodic time fluctuation quantity of the high frequency component quantity D1. Here, FIG. 1 is a block diagram showing the high-frequency component variation feature quantity measuring unit 4. In the high-frequency component variation feature quantity measuring unit 4, the bandpass filter (BPF) unit 11 corresponds to the B picture of the input video signal S1 and the non-picture corresponding to the periodic pattern of the picture type set in the input video signal S1. A frequency component corresponding to the periodicity of the B picture is extracted and output. Specifically, when the GOP structure of the input video signal S1 is the GOP structure shown in FIG. 3, the band-pass filter unit 11 has a 3-tap FIR with tap coefficients of 2, -1, and -1 as shown in FIG. A periodic fluctuation component having a three frame period, which is a repetition period of the I and P pictures, is extracted from the high frequency component amount D1 by the filter. In this case, as shown in FIG. 5 by comparison with FIG. 4, the bandpass filter unit 11 may be further narrowed by setting the number of taps to 3 taps or more, for example, 6 taps. Further, a tap coefficient may be set as shown in FIG. 6 in comparison with FIGS. 4 and 5 to reflect the difference in the high frequency component amount D1 between the I picture and the P picture in the extraction result.

  As shown in FIG. 7, the high-pass filter (HPF) unit 12 sets a frequency band including the pass band L11 of the band-pass filter unit 11 as a pass band, and extracts a signal component of the pass band from the high-frequency component amount D1. The output signal D4 is output. Here, FIG. 8 is a block diagram showing a configuration of the high-pass filter unit 12. In the high pass filter unit 12, the average value filter unit 14 is an average value filter having the same tap length as the band pass filter unit 11, and averages the high frequency component amount D1 to obtain the DC value of the high frequency component amount D1. Output. Therefore, for example, when the bandpass filter unit 11 is configured with 6 taps according to the characteristics shown in FIG. 5, the average value filter unit 14 is applied with an average value filter having tap coefficients of 1, 1, 1, 1, 1, 1. The

  The high pass filter unit 12 subtracts the output signal of the average value filter unit 14 from the high frequency component amount D1, and outputs an output signal D4. The high-pass filter unit 12 is not limited to the configuration shown in FIG. 8, and various configurations can be widely applied.

  The bandpass filter (BPF) output delay unit 16 sequentially delays the output signal D3 of the bandpass filter unit 11 by one sampling period and outputs the delayed signal by a plurality of systems. The high-pass filter (HPF) output delay unit 17, in the same way as the band-pass filter output delay unit 16, sequentially delays the output signal D 4 of the high-pass filter unit 12 by a predetermined time and outputs it by a plurality of systems. Thereby, the band pass filter output delay unit 16 and the high pass filter output delay unit 17 output a plurality of output signals D3 and outputs from the band pass filter unit 11 and the high pass filter unit 12 within a certain period determined by the maximum delay time. The signals D4 are output simultaneously and in parallel.

  The bandpass filter (BPF) output energy calculation unit 18 receives the output signals D3 of a plurality of systems output from the bandpass filter output delay unit 16 and the output signal D3 output from the bandpass filter unit 11, and outputs these signals. The energy value of the signal D3 is calculated. For this energy calculation, various calculation methods such as sum of squares, square root of sum of squares, sum of absolute values, and the like can be applied.

  A high-pass filter (HPF) output energy calculation unit 19 receives a plurality of output signals D4 output from the high-pass filter output delay unit 17 and an output signal D4 output from the high-pass filter unit 12, and calculates a band-pass filter output energy. Similar to the unit 18, the energy values of these output signals D4 are calculated.

  The fluctuation period component occupancy calculating unit 20 divides the energy value D7 calculated by the bandpass filter output energy calculating unit 18 by the energy value D9 calculated by the highpass filter output energy calculating unit 19, and thereby the input video signal The fluctuation cycle component occupancy D10, which is the ratio of the cycle fluctuation component in the high frequency component of S1, is detected. Therefore, in this case, the larger the value of the fluctuation period component occupancy D10, the greater the proportion of the high frequency component of the input video signal S1 occupied by the period fluctuation component of the picture type.

  The fluctuation cycle stability calculation unit 21 receives the fluctuation cycle component occupancy D10 calculated by the fluctuation cycle component occupancy calculation unit 20, and the fluctuation cycle component occupancy as shown in FIGS. The fluctuation cycle stability D11 indicating the stability in the time axis direction of the degree D10 is calculated. Here, the fluctuation cycle stability D11 can be obtained by calculating the variance, standard deviation, average deviation, etc. of the fluctuation cycle component occupancy D10. In the example shown in FIG. 9, it can be seen that during the period T, the change in the fluctuation period component occupancy D10 is smaller than that in the other periods, and the fluctuation period stability D11 increases in this period T.

  The variation feature value generation unit 23 corrects the energy value D7 calculated by the bandpass filter output energy calculation unit 18 with the variation period component occupancy D10 and the variation period stability D11, and the picture set in the input video signal S1. A variation feature amount D12 indicating the magnitude of the periodic variation component of the high frequency component depending on the type is output.

  Here, FIG. 10 is a block diagram showing a detailed configuration of the fluctuation feature quantity generation unit 23. In the fluctuation feature value generation unit 23, the occupancy degree reliability setting unit 25 determines the fluctuation period component occupancy D10 based on predetermined thresholds Rth1 and Rth2, and the fluctuation period component occupancy D10 is a threshold Rth1 as shown in FIG. The value is 0 in the following cases, the value is 1 when the fluctuation period component occupancy D10 is equal to or greater than the threshold value Rth2, and the fluctuation period when the fluctuation period component occupancy D10 is equal to or greater than the threshold value Rth1 and less than or equal to the threshold value Rth2. The occupancy reliability D13 is generated so that the value increases linearly as the component occupancy D10 increases. Here, the occupancy reliability D13 is a parameter indicating the probability that the fluctuation period component occupancy D10 reflects the period fluctuation due to the picture type.

  The stability reliability setting unit 26 determines the fluctuation cycle stability D11 based on the predetermined threshold values Sth1 and Sth2, and as shown in FIG. 12, when the fluctuation cycle stability D11 is equal to or less than the threshold value Sth1, the value is 0. When the fluctuation cycle stability D11 is greater than or equal to the threshold value Sth2, the value is 1. When the fluctuation cycle stability D11 is greater than or equal to the threshold value Sth1 and less than or equal to the threshold value Sth2, the value increases linearly as the fluctuation cycle stability D11 increases. As described above, the stability reliability D14 is generated. Here, the stability reliability D14 is a parameter indicating the probability that the fluctuation period stability D11 reflects the period fluctuation due to the picture type.

  The variation feature reliability setting unit 27 multiplies the occupancy reliability D13 and the stability reliability D14, and the energy value D7 calculated by the bandpass filter output energy calculation unit 18 reflects the periodic variation depending on the picture type. Fluctuation feature reliability D15, which is a certainty, is calculated.

  The variation feature amount calculation unit 28 determines the variation feature reliability D15 with a predetermined threshold, and based on the determination result, when the variation feature reliability D15 is equal to or greater than the threshold, the bandpass filter output energy calculation unit 18 The calculated energy value D7 is output as it is as the variation feature amount D12. On the other hand, when the variation feature reliability D15 is smaller than this threshold, the variation feature amount D12 is set to a value of 0 and output.

  Through these processes, the high-frequency component variation feature amount measurement unit 4 evaluates the degree of concentration on a specific frequency related to the periodic variation by the picture type and its temporal stability, and generates the reliability levels D13 and D14. Based on the degrees D13 and D14, the fluctuation feature quantity D12 indicating the magnitude of the cyclic fluctuation component of the high frequency component quantity D1 is generated.

  The variation feature amount time smoothing unit 30 (FIG. 1) smoothes the variation feature amount D12 detected by the variation feature amount generation unit 23, and outputs a high frequency component variation feature amount D2. For the smoothing process, a low pass filter, a median filter, or the like having an IIR filter configuration or an FIR filter configuration can be applied.

  FIG. 13 is a block diagram showing the video signal processing unit 2. In the video signal processing unit 2, the noise reduction processing unit 32 suppresses the noise of the input video signal S1 by suppressing the random noise components in the image space direction and the time axis direction included in the input video signal S1, and outputs the result. . The sharpness enhancement processing unit 33 emphasizes the specific frequency component of the video signal S3 output from the noise reduction processing unit 32, thereby enhancing and outputting the sharpness of the input video signal S1. The subsequent contrast enhancement processing unit 34 enhances the contrast according to the luminance distribution in the screen of the video signal S4 output from the sharpness enhancement processing unit 33, and outputs it. Specifically, the contrast enhancement processing unit 34 amplifies the input video signal S1 with nonlinear characteristics to enhance contrast, as indicated by reference numeral L34 in FIG. 14, and this nonlinear characteristic is converted into the average luminance level in the screen. And dynamically changing based on a luminance histogram or the like. FIG. 14 shows nonlinear characteristics for emphasizing the intermediate gradation. In this case, the amplification factor decreases on the higher and lower signal levels of the input video signal S1, and the amplification factor of the intermediate gradation is correspondingly reduced. Is set to increase. When emphasizing the dark gradation, the contrast enhancement processing unit 34 increases the amplification factor on the side where the signal level of the input video signal S1 is low, and accordingly reduces the amplification factor on the intermediate gradation and high gradation. . The video signal processing apparatus 1 of this embodiment outputs a video signal in which the contrast is enhanced by the contrast enhancement processing unit 34 as an output video signal S2.

  In the video signal processing, any one or two of these noise reduction processing, sharpness enhancement processing, and contrast enhancement processing may be executed, and further combined with other processing. You may make it perform. Further, the order of the processes may be changed, and further, these processes may be executed simultaneously and in parallel, and a plurality of video signals obtained as a result may be synthesized to be the output video signal S2.

  The control information generation unit 35 controls the operations of the noise reduction processing unit 32, the sharpness enhancement processing unit 33, and the contrast enhancement processing unit 34 based on the high frequency component amount D1 and the high frequency component variation feature amount D2, respectively. D16, D17, and D18 are output.

  Here, the high-frequency component amount D1 detected by the high-frequency component measuring unit 3 detects a component having the property of random noise. As a result, the control information generation unit 35 controls the noise reduction processing unit 32 so that the noise suppression effect increases as the value of the high frequency component amount D1 increases, and thereby the noise component included in the output signal S3 is time-based. The noise in the input video signal S1 is reduced by making it uniform in the direction. In addition, in this embodiment, the noise component is made uniform in the time axis direction in the present embodiment, and the time fluctuation of the high frequency component related to the flicker deterioration is also made uniform in the time axis direction.

  On the other hand, when the high frequency component variation feature amount D2 shows a large value, it indicates that the temporal variation of the high frequency component amount due to the picture type has occurred and the variation range is large. Will be. As a result, the control information generation unit 35 determines the value of the high-frequency component variation feature value D2 with a predetermined threshold value, and when the value of the high-frequency component variation feature value D2 is larger than this threshold value, the noise reduction effect increases more than usual. Thus, the noise reduction processing unit 32 is controlled. In addition, when it is more than this threshold, it may be set so that the noise reduction effect is gradually increased by the increase of the high frequency component variation feature amount D2, and the noise is reduced step by step by the increase of the high frequency component variation feature amount D2. You may set so that an effect may increase. Further, the noise reduction effect may be switched in two steps depending on whether the threshold is greater than or equal to the threshold and less than the threshold.

  Thereby, the control information generation unit 35 suppresses the fluctuation of the noise component in the time axis direction according to the high frequency component amount D1, and suppresses the noise component in the time axis direction by the high frequency component fluctuation feature amount D2. Variable, reduces flicker degradation and suppresses noise components.

  On the other hand, the sharpness enhancement processing in the sharpness enhancement processing unit 33 is processing that emphasizes the specific frequency component and emphasizes the sharpness, and therefore the control information generation unit 35 determines that the value of the high frequency component amount D1 is The sharpness enhancement processing unit 33 is controlled so that the amplification factor of the specific frequency band related to the enhancement of sharpness decreases as it increases, thereby uniformizing the high frequency component included in the output signal S4 in the time axis direction. . In addition, the control information generation unit 35 determines the value of the high-frequency component variation feature value D2 with a predetermined threshold, and when the value of the high-frequency component variation feature value D2 is larger than the threshold value, the specific frequency related to the enhancement of the sharpness The sharpness enhancement processing unit 33 is controlled so that the amplification factor of the band is lowered. In addition, when it is more than this threshold, it may be set so that the amplification factor gradually decreases with the increase of the high frequency component variation feature amount D2, and the amplification factor is gradually increased by the increase of the high frequency component variation feature amount D2. You may set so that it may fall. Further, the amplification factor may be switched in two steps depending on whether it is equal to or greater than the threshold and less than the threshold. Thereby, the control information generation unit 35 reduces the flicker deterioration by reducing the width of the periodic time fluctuation of the high frequency component amount included in the output signal S4 of the sharpness enhancement processing unit 33.

  Further, the control information generation unit 35 determines the value of the high-frequency component variation feature value D2 with a predetermined threshold, and when the value of the high-frequency component variation feature value D2 is larger than the threshold value, the increase of the high-frequency component variation feature value D2 Accordingly, the non-linear characteristic related to contrast enhancement is weakened, and the contrast enhancement processing unit 34 is controlled so as to approach the linear characteristic indicated by the symbol L34A in FIG. Note that even when the threshold value is greater than or equal to this threshold, the linear characteristic may be gradually set closer to an increase in the high-frequency component variation feature amount D2, or the linear characteristic is gradually increased as the high-frequency component variation feature amount D2 increases. You may set so that it may approach a characteristic. The characteristics may be switched in two steps depending on whether the threshold is equal to or greater than the threshold and less than the threshold. As a result, the control information generation unit 35 reduces flicker degradation in the output video signal S2.

  In this way, in this embodiment, among the video signal processing of the input video signal S1 in the noise reduction processing unit 32, the sharpness enhancement processing unit 33, and the contrast enhancement processing unit 34, only the contrast enhancement processing in the contrast enhancement processing unit 34, Although the case where the characteristics are switched according to the input video signal S1 and the input video signal S1 is processed with nonlinear characteristics has been described, the present invention is not limited to this, and the noise reduction processing section 32 and the sharpness enhancement processing section. The processing at 33 can also be widely applied when the characteristics are switched according to the input video signal S1 and the input video signal S1 is processed with non-linear characteristics. Needless to say, even when the input video signal S1 is processed with such a fixed characteristic as in the noise reduction processing unit 32, flicker deterioration can be reduced.

(2) Operation of Embodiment In the above configuration, the input video signal S1 (FIG. 2) is subjected to noise reduction processing, sharpness enhancement processing, and contrast enhancement processing by the video signal processing unit 2 to improve the image quality, and the output video signal Output by S2. Here, since these noise reduction processing, sharpness enhancement processing, and contrast enhancement processing are processing for high-frequency components of the input video signal S1, the input video signal S1 is a video signal transmitted after being compressed by MPEG or the like. In some cases, the periodic fluctuation (FIG. 3) of the high frequency component due to the picture type set in the input video signal S1 is perceived as flicker degradation when the output video signal S2 is displayed.

  Therefore, in the video signal processing device 1, the high frequency component amount D1 of the input video signal S1 is detected by the high frequency component measuring unit 3, and the high frequency component variation feature amount measuring unit 4 processes the high frequency component amount D 1. A high frequency component variation feature amount D2 indicating a periodic time variation amount of the high frequency component amount D1 is detected. Further, the fluctuation in the time axis direction of the noise reduction processing is suppressed by the high frequency component amount D1, and the periodic variation in the high frequency component of the output video signal S2 is suppressed by the high frequency component variation feature amount D2. In addition, noise reduction processing, sharpness enhancement processing, and contrast enhancement processing are variably controlled. As a result, in this video signal processing apparatus, it is possible to change the characteristics of the video signal processing according to the magnitude of the periodic fluctuation of the high frequency component related to the flicker deterioration so that the flicker deterioration is not noticeable. In comparison, the flicker deterioration can be reduced. Further, by suppressing the periodic fluctuation in the high frequency component of the output video signal S2 by changing the characteristics, the characteristics of the video signal processing can be reliably changed even when the characteristics of the video signal processing are adaptively changed according to the input video signal S1. Flicker degradation can be reduced.

  That is, the input video signal S1 is detected by the high frequency component measuring unit 3 by separating the high frequency component in the image space direction and / or the high frequency component in the time direction, and detecting the high frequency component amount. By detecting the high frequency component amount in the direction and / or the high frequency component in the time direction, it is possible to reliably reduce the flicker deterioration. Further, the noise suppression effect can be controlled by using this high frequency component amount for noise suppression processing, and the overall configuration can be simplified. That is, according to the high frequency component in the image space direction, for example, even if the high frequency component amount detection unit is set to a frame and the high frequency component amount changes at various points in one screen, It is possible to reliably detect the component amount of the high frequency component. Further, according to the high frequency component in the time direction, even when the component amount of the high frequency component fluctuates with time, the component amount of the high frequency component that causes flicker degradation can be reliably detected.

  Further, in the processing in the high frequency component measuring unit 3, the high frequency component in the image space direction having a specific property in the time direction is separated and the high frequency component amount is detected, thereby increasing the high frequency component in the image space direction and the time direction. The high-frequency component amount can be detected in the same manner as the detection of the high-frequency component amount, thereby simplifying the overall configuration and reliably detecting the component amount of the high-frequency component that causes flicker degradation. In addition, by applying the random noise property to the specific property in the time direction, the entire configuration can be simplified by using this high frequency component amount for noise suppression processing.

  On the other hand, in the high-frequency component variation feature amount measuring unit 4, the high-frequency component amount D1 of the input video signal S1 is input to the bandpass filter unit 11 (FIG. 1), and is set to the input video signal S1 here. The repeated periodic components of the B picture and the non-B picture corresponding to the picture type periodic pattern are extracted (FIGS. 4 to 6). Further, the high frequency component variation feature quantity D2 is generated by the repetition period component. As a result, the video signal processing apparatus 1 can reliably reduce the flicker deterioration resulting from the GOP structure.

  More specifically, the high frequency component of the high frequency component amount D1 is detected in the input video signal S1 by the high pass filter unit 12 in which the pass band is set so as to include the pass band of the band pass filter unit 11 (FIG. 7). And FIG. 8), it is possible to detect the ratio of the periodic fluctuation component in the high frequency component of the input video signal S1. The input video signal is processed by these output signals so as to be corrected by the output signal of the high-pass filter unit 12 based on the output signal of the band-pass filter unit 11, thereby generating a high-frequency component variation feature amount D2. . As a result, the video signal processing apparatus 1 can prevent the flicker degradation reduction process from malfunctioning due to, for example, the amount of high-frequency components inherently included in the input video signal S1, thereby reliably preventing flicker degradation. It becomes possible to reduce.

  That is, in the video signal processing device 1, the energy values D7 and D9 of the output signal of the bandpass filter unit 11 and the output signal of the highpass filter unit 12 are obtained by the bandpass filter output energy calculation unit 18 and the highpass filter output energy calculation unit 19, respectively. Then, the energy value D7 is divided by the energy value D9 to obtain the fluctuation period component occupancy D10 which is a ratio of the period fluctuation component to the high frequency component of the input video signal S1. Further, the fluctuation cycle stability D11 indicating the stability of the fluctuation cycle component occupancy D10 in the time axis direction is obtained by the fluctuation cycle stability calculation unit 21 (FIG. 9). Further, the energy value D7 of the output signal of the bandpass filter unit 11 is corrected by the fluctuation period component occupancy D10 and the fluctuation period stability D11 to obtain the high frequency component fluctuation feature amount D2.

  As a result, in the video signal processing apparatus 1, even if a change in signal level similar to the picture type periodic variation component occurs temporarily in the input video signal S1, the high frequency range of the input video signal S1 is further increased. Even when the components are variously changed, the flicker deterioration can be reliably suppressed.

  That is, in the input video signal S1, the fluctuation period component occupancy D10 and the fluctuation period stability D11 are determined by the predetermined threshold values Rth1, Rth2, Sth1, and Sth2, respectively (FIGS. 10, 11, and 12). When D10 and fluctuation cycle stability D11 are extremely small, it is determined that the reliability D13 and D14 are low. Further, the reliability D13 and D14 are set to increase and saturate at a value of 1 as the fluctuation cycle component occupancy D10 and the fluctuation cycle stability D11 increase. Further, a multiplication value of the reliability levels D13 and D14 is obtained, thereby obtaining an overall reliability level.

  When the total reliability of the input video signal S1 is equal to or greater than a certain value, the energy value D7 of the output signal of the bandpass filter unit 11 is output as the fluctuation feature quantity D12, and the fluctuation feature quantity D12 is smoothed. Thus, the high-frequency component variation feature amount D2 is obtained.

  The input video signal S1 (FIG. 13) is sequentially subjected to noise reduction processing, sharpness enhancement processing, and contrast enhancement processing by the noise reduction processing unit 32, the sharpness enhancement processing unit 33, and the contrast enhancement processing unit 34 of the video signal processing unit 2. Is output by the output video signal S2. Also, in these processes, in the contrast enhancement process, the contrast is amplified by the non-linear characteristic, and the non-linear characteristic is dynamically changed based on the average luminance level in the screen of the input video signal S1, the luminance histogram, and the like. By being changed, the nonlinear characteristic is dynamically changed according to the input video signal S1, and the contrast of the input video signal S1 is reduced.

  As a result, with respect to the contrast enhancement processing by the contrast enhancement processing unit 34, the input video signal S1 is detected so that the luminance signal of the input video signal is simply gamma-corrected to detect the average luminance level difference between frames and this level difference is suppressed. If the signal level is corrected, it is difficult to reduce the flicker deterioration.

  However, if the periodic variation component of the high frequency component is detected and the periodic variation component is set to be suppressed by changing the characteristic related to contrast enhancement as in this embodiment, the input video signal S1 can be variously displayed by nonlinear characteristics. Even in the case of processing the same, flicker degradation can be reliably reduced.

  Thus, in this embodiment, the noise reduction processing unit 32 is controlled so that the noise suppression effect increases as the value of the high frequency component amount D1 increases, and thereby the noise component included in the output signal S3 is changed in the time axis direction. The noise of the input video signal S1 is reduced so as to be uniform, and the time variation of the high frequency component related to the flicker deterioration is further reduced. As a result, in this embodiment, the configuration of the noise reduction processing unit 32 is simplified by performing the measurement of the amount of noise related to the noise reduction processing by the high frequency component measurement unit 3. Also, the value of the high-frequency component variation feature value D2 is determined with a predetermined threshold, and if the value of the high-frequency component variation feature value D2 is larger than this threshold value, the noise reduction effect is increased more than usual, and thus perceived by noise reduction processing Reduces possible flicker degradation.

  Further, the sharpness enhancement processing unit 33 is controlled so that the amplification factor of the specific frequency band related to sharpness enhancement decreases as the value of the high frequency component amount D1 increases, and thereby the high frequency component included in the output signal S4. In the time axis direction. Further, when the value of the high-frequency component variation feature value D2 is determined with a predetermined threshold value and the value of the high-frequency component variation feature value D2 is larger than the threshold value, the amplification factor of the specific frequency band related to the sharpness enhancement is reduced. As described above, the sharpness enhancement process is controlled to reduce the flicker deterioration that can be perceived by the sharpness enhancement process.

  Further, when the value of the high-frequency component variation feature quantity D2 is determined with a predetermined threshold value and the value of the high-frequency component variation feature value D2 is larger than this threshold value, the non-contrast enhancement is performed according to the increase of the high-frequency component variation feature value D2. The contrast enhancement process is controlled so that the linear characteristic is weakened to approach the linear characteristic (FIG. 14). This reduces flicker degradation that can be perceived by contrast enhancement processing.

(3) Effects of the embodiment According to the above configuration, the periodic time fluctuation amount of the high frequency component amount included in the input video signal is detected, and the periodic time fluctuation is suppressed based on the detection result. Thus, by varying the characteristics of the video signal processing, it is possible to reduce the flicker deterioration more reliably than in the past.

  Further, by detecting the high frequency component amount by separating the high frequency component in the image space direction and / or the high frequency component in the time direction from the input video signal, it is possible to reliably reduce flicker degradation.

  In addition, by separating the high frequency component in the image space direction having specific properties in the time direction and detecting the high frequency component amount, the high frequency component is detected in the same manner as the high frequency component amount is detected in the image space direction and the time direction. The amount of components can be detected, whereby the overall configuration can be simplified and the amount of high-frequency components that cause flicker degradation can be reliably detected. In addition, by applying the random noise property to the specific property in the time direction, the entire configuration can be simplified by using this high frequency component amount for noise suppression processing.

  Further, by detecting the periodic temporal fluctuation amount of the high frequency component amount corresponding to the picture type periodic pattern set in the input video signal, it is possible to reliably reduce the flicker deterioration derived from the GOP structure.

  In addition, a signal component corresponding to this periodic pattern is extracted from the amount of the high frequency component, a high frequency component including this signal component is extracted, and the signal component related to this periodic pattern is corrected with the extracted high frequency component to obtain a high frequency component. By generating the component variation feature amount, it is possible to prevent the malfunction of the flicker degradation reduction process due to the amount of the high frequency component inherently included in the input video signal S1, thereby reliably reducing the flicker degradation. It becomes possible to do.

  That is, calculate the ratio by calculating the energy of the signal component due to the bandpass output and the highpass output respectively, calculate the fluctuation period stability indicating the stability of the fluctuation period component occupancy that is this ratio in the time axis direction, and the bandpass output By detecting the high frequency component fluctuation feature quantity by correcting the energy value of the fluctuation period component occupancy and fluctuation period stability, a signal level change similar to that of the picture type periodic fluctuation component occurred temporarily. Even in this case, even when the high frequency component of the input video signal changes variously, it is possible to reliably suppress flicker degradation.

  Also, when the characteristics are varied according to the input video signal and the input video signal is amplified by the non-linear characteristics to enhance the contrast, the non-linear characteristics are converted into the linear characteristics based on the high-frequency component variation feature quantity. By suppressing the periodic fluctuations close to, flicker degradation can be reliably suppressed when the characteristics are varied according to the input video signal to enhance the contrast of the video signal.

  FIG. 15 is a block diagram illustrating a configuration of a high-frequency component variation feature quantity measurement unit applied to the video signal processing apparatus according to the second embodiment of the present invention. The video signal processing apparatus according to the second embodiment is configured in the same manner as the video signal processing apparatus 1 according to the first embodiment except that the configuration of the high-frequency component variation feature quantity measuring unit 44 is different. The high-pass component variation feature quantity measurement unit 44 omits the high-pass filter unit 12 and replaces the high-pass filter output delay unit 17, the band-pass filter output energy calculation unit 18, and the high-pass filter output energy calculation unit 19 with a high-pass component amount. Except for the provision of a delay unit 46, a bandpass filter (BPF) output energy calculation unit 48, and a high-frequency component energy calculation unit 49, the configuration is the same as the high-frequency component variation feature value measurement unit of the first embodiment.

  Here, the high-frequency component amount delay unit 46 directly inputs the high-frequency component amount D1, and in the same way as the bandpass filter output delay unit 16, delays the high-frequency component amount D1 sequentially by one sampling period by a plurality of systems. Output.

  The band-pass filter output energy calculation unit 48 receives a plurality of output signals D6 output from the band-pass filter output delay unit 16 and an output signal D3 output from the band-pass filter unit 11, and outputs the maximum value by detecting the maximum value. An energy value D7 of the signal D3 is calculated.

  That is, FIG. 16 is a block diagram showing the bandpass filter output energy calculation unit 48. In the band pass filter output energy calculation unit 48, the maximum value detection unit 51 receives a plurality of output signals D 6 output from the band pass filter output delay unit 16 and an output signal D 3 output from the band pass filter unit 11. To do. The maximum value detector 51 detects and outputs the maximum value D22 from these output signals D6 and D3. The timing at which the maximum value D22 is detected from the output signal D3 is output to the high frequency component energy calculation unit 49 as the maximum value detection synchronization signal D21.

  That is, on the premise of the GOP structure described above with reference to FIG. 3, when the bandpass filter unit 11 is created with the 6 taps described above with reference to FIG. 5, the base waveform based on the filter coefficients of this bandpass filter unit 11 is 3 as shown in FIG. It will stand up with a frame period. Accordingly, in the output signal D3 of the bandpass filter unit 11, when the input video signal S1 includes a periodic variation component depending on the picture type, a change in signal level corresponding to the repetition of the base waveform occurs as shown in FIG. It will be.

  Here, when the five output signals D6 are output from the bandpass filter output delay unit 16 corresponding to the configuration of the bandpass filter unit 11, the bandpass filter output energy calculation unit 48 uses the sampling value B [t ] Is input by the output signal D3 from the band-pass filter output delay unit 16 at five sampling times t-1, t-2, t-3, t-4, t-5 immediately before this time t. Sampling values B [t-1], B [t-2], B [t-3], B [t-4], and B [t-5] are input by the output signal D6. Thus, the maximum value detection unit 51 detects the maximum sampling value B [t] from these six sampling values B [t] to B [t-5] and outputs the maximum value D22 at this time t. .

  The normal circuit energy calculation unit 52 detects the normalized energy value from the maximum value D22 output from the maximum value detection unit 51, and uses this normalized energy value as the energy of the output signal D3 output from the bandpass filter unit 11. Output as value D7. Here, the normal circuit energy calculation unit 52 normalizes by dividing the maximum value D22 by the norm of the basis vector based on the filter coefficients [2, -1, -1, 2, -1, -1] of the bandpass filter unit 11. Calculate the energy value. The norm of the basis vector is obtained by the square root of the sum of squares of each element of the basis vector.

  The high frequency component energy calculation unit 49 corresponds to the processing of the band pass filter output energy calculation unit 48, so that the high frequency component amount delay unit 46 outputs a plurality of output signals D23, the high frequency component measurement unit 3 The output high frequency component amount D1 is processed to calculate the high frequency component energy value D9.

  Here, FIG. 19 is a block diagram showing a configuration of the time AC energy calculation unit 49. In the time AC energy calculation unit 49, the synchronization data detection unit 54 delays the high frequency component amount at the timing when the maximum value is detected in the output signal D3 of the bandpass filter unit 11 with reference to the maximum value detection synchronization signal D21. A plurality of output signals D23 output from the unit 46 and the high frequency component amount D1 output from the high frequency component measuring unit 3 are acquired and output. As a result, the synchronization data detection unit 54 outputs the sampling values R [t] to R [t-5] in the example of FIG.

The normalized average value filter unit 55 receives the output data D24 of the synchronization data detection unit 54, and performs normalized averaging on the output data D24. Here, the normalization process is executed by dividing the output data D24 by the norm of the basis vector having the same filter coefficient as an element. Thus, in this embodiment, by dividing six sample value R [t] ~R [t-5] with each ^61/2 executes the process of normalization. The normalized average value filter unit 55 adds up the normalized sampling values, thereby detecting and outputting the DC value of the high frequency component amount D1.

  The high frequency component amount time DC energy calculation unit 56 calculates and outputs the square value of the output data D25 of the normalized average value filter unit 55. The high frequency component amount time energy calculating unit 57 calculates the square value of the output data D24 (R [t] to R [t-5]) of the synchronous data detecting unit 54, and thereby the time of the high frequency component amount D1. The direction energy value D27 is output.

  The subtraction circuit 58 subtracts the output data D26 of the high frequency component amount time DC energy calculation unit 56 from the output data D27 of the high frequency component amount time energy calculation unit 57, and obtains the energy value D9 of the AC component of the high frequency component amount D1. calculate.

  As a result, the high-frequency component variation feature amount measurement unit 44 normalizes and detects the maximum value energy from the bandpass filter within a specific period, and also uses the maximum-value timing as a reference for the time DC energy of the high-frequency component amount. Is normalized, and the normalized maximum time DC energy is subtracted from the normalized high frequency component time energy to output the high frequency component amount temporal AC component energy value D9.

Here, as shown in FIG. 20, it is assumed that these energies are expressed in an orthonormal coordinate space. Here, the orthonormal coordinate space is a six-dimensional space because the number of filter taps used in the bandpass filter unit 11 and the normalized average value filter unit 55 is six. A high-frequency component amount D1 that is an input of the bandpass filter unit 11 and the normalized average value filter unit 55 is converted into a six-dimensional vector Y having six consecutive sampling values as elements in this orthonormal coordinate space. To express. In this case, the output signal D25 of the normalized average value filter unit 55 becomes the norm of the vector X _DC obtained by projecting the vector Y onto the DC axis, and this output signal D25 is processed by the high frequency component amount time DC energy calculation unit 56. energy value D26 obtained Te will correspond to the square value of the norm of the vector X _DC. On the other hand, the energy value D27 output from the high frequency component amount time energy calculating unit 57 corresponds to the square value of the norm of the six-dimensional vector Y.

Further, the energy value D7 obtained by the bandpass filter output energy calculation unit 18 is 2 of the norm of the vector X _BP which is a vector Y projected on the axis in the vector direction whose elements are the coefficients of the filters constituting the bandpass filter unit 11. It corresponds to a multiplier value. Energy values D9 obtained by subtracting the energy value D26 from the energy value D27 high frequency component amount time DC energy calculating unit 56 of the high-frequency component amount time energy calculating unit 57 on the other hand, a plane perpendicular to the vector X _DC This corresponds to the square value of the norm of the vector X _AC which is projected onto the vector Y. Here, the vector X _ETC in FIG. 20 represents an AC component other than the vector X _BP , and can be regarded as a combined vector of four vectors corresponding to the remaining AC components.

Thus the ratio of the norm and the vector X _AC of the norm of the vector X _BP, or by calculating the square value of this ratio, in the same manner as in Example 1, it can be seen that can detect variation period component occupancy D10.

  In this video signal processing apparatus, the fluctuation period component occupancy calculation unit 50 detects the fluctuation period component occupancy D10 by dividing the energy value D7 by the energy value D9 in the same manner as described above for the first embodiment.

  Thereby, in this embodiment, the high frequency component amount D1 is expressed by a vector, normalized by a base vector to obtain an energy value, and a periodic time variation amount of the high frequency component amount included in the input video signal is detected. Thus, the same effect as in the first embodiment can be obtained.

  Further, by omitting the high-pass filter unit 12 and directly processing the high-frequency component amount obtained by the high-frequency component measuring unit to detect the high-frequency component variation feature value, the same configuration as that of the first embodiment can be obtained. An effect can be obtained.

  FIG. 21 is a block diagram illustrating a configuration of a high-frequency component variation feature amount measurement unit applied to the video signal processing device according to the third embodiment of the present invention. The video signal processing apparatus according to the third embodiment is configured in the same manner as the video signal processing apparatus 1 according to the first embodiment except that the configuration of the high frequency component variation feature amount measuring unit 64 is different.

  In the high frequency component variation feature amount measuring unit 64, the AC component vector component unit 65 vectorizes and outputs the time variation component of the high frequency component amount D1. That is, as shown in FIG. 22, in the AC component vector generation unit 65, the high frequency component amount delay unit 66 sequentially delays the high frequency component amount D <b> 1 from the number of input taps of the normalized average value filter unit 67 in the subsequent stage. Output by multiple systems with a value of 1 less.

  The vector data generation unit 68 receives the output data D23 and the high frequency component amount D1 of the high frequency component amount delay unit 66, and generates vector data DV1 having the output data D23 and the high frequency component amount D1 as elements, respectively. Output. Accordingly, the vector data DV1 is vector data having the same number of dimensions as the number of input taps of the subsequent normalized average value filter unit 67, and indicates a signal waveform based on consecutive sampling values of the high frequency component amount D1.

  The normalized average value filter unit 67 outputs from the vector data generation unit 68 by a filter in which all the filter coefficients are the same and the coefficients are set so that the size of the vector having the filter coefficient as an element is 1. The vector data DV1 is subjected to a convolution operation and an operation result D25 is output. Here, the calculation result D25 is equivalent to the inner product calculation result of the vector data DV1 and a vector composed of the filter coefficient of the normalized average value filter unit 67. Thus, the normalized average value filter unit 67 detects the DC value D25 of the high frequency component amount D1.

  The DC component vector generation unit 69 scales the vector composed of the filter coefficients of the normalized average value filter unit 67 by the calculation result D25 output from the normalized average value filter unit 67, and the DC value D25 of the high frequency component amount D1. Is output by the DC component vector DV3 corresponding to the vector data DV1.

The subtraction circuit 70 subtracts the DC component vector DV3 from the vector data DV1 output from the vector data generation unit 68, and outputs an AC component vector DV4 obtained by vectorizing the time variation component of the high frequency component amount D1. Here, in the orthonormal coordinate space of FIG. 20, the vector data DV1, the DC component vector DV3, and the AC component vector DV4 correspond to the vector Y, the vector X _DC , and the vector X _AC , respectively.

  The band-pass vector phase correlation calculating unit 73 (FIG. 21) processes the AC component vector DV4 to indicate a normalized band-pass inner product indicating the magnitude of periodic time fluctuation of the high-frequency component amount D1 related to flicker degradation. A band pass vector phase correlation degree D34 representing the magnitude of this periodic time fluctuation component with respect to all the time fluctuation components of the signal D32 and the high frequency component amount D1 is detected.

That is, as shown in FIG. 23, in the bandpass vector phase correlation calculation unit 73, the normalized bandpass vector inner product calculation unit 75 calculates the normalized bandpass vector E _BP and the AC component vector DV4 (X _AC ). Calculate inner product. The normalized band pass vector E _BP has a tap coefficient that is an AC component with respect to the filter coefficient of the band pass filter that is set by modeling the periodic time variation of the high frequency component amount D1 related to flicker degradation. The vector is normalized so that the size of the vector having the same number of dimensions as the vector DV4 and the filter coefficient as an element is 1. Note that the bandpass filter set by modeling the periodic temporal fluctuation of the high frequency component amount D1 related to the flicker degradation is, for example, the bandpass filter unit 11 in FIG. As a result, the bandpass vector phase correlation calculation unit 73 detects the magnitude of the periodic time fluctuation of the high frequency component amount D1 related to flicker degradation from the inner product value obtained by the inner product calculation, and normalizes the bandpass inner product signal D32. Output as.

  The AC component vector norm calculation unit 76 calculates and outputs the norm D33 of the AC component vector DV4.

The vector phase correlation degree calculation unit 77 detects the bandpass vector phase correlation degree D34 (cos θ) by an arithmetic process represented by the following equation. Here, <X _AC , E _BP > is the normalized bandpass inner product signal D32, and the absolute value of X _AC is the norm D33 of the AC component vector DV4.

Here, as shown in FIG. 24, in the orthonormal coordinate space, the bandpass vector E _BP is expressed by E _BP = X _AC / | X _AC |. increasing period fluctuation component, as the angle between AC component vector DV4 and (X _AC) and band-pass vector E _BP theta is the smaller, (1) a band-pass vector phase correlation according to equation D34 value is greater Become. That is, as the bandpass vector phase correlation degree D34 is larger, the direction of the AC component vector DV4 indicating the time variation of the high frequency component amount D1 is closer to the bandpass vector indicating the periodic time variation. . Note that <X _AC , E _BP > in the equation (1) becomes | X _AC || E _BP | cos θ = | X _AC | cos θ = | X _BP |, and the normalized bandpass inner product signal D32 <X _It can be seen that _AC , E _BP > is the norm of the vector X _BP obtained by projecting the vector Y onto the axis in the direction of the normalized bandpass vector E _BP on the orthonormal space shown in FIG. Therefore, it can be seen that cos θ = | X _BP | / | X _AC |, and the index of the periodic fluctuation component detected in the above-described second embodiment is detected in the same manner.

  As a result, the vector phase correlation calculating unit 77 calculates the entire high-frequency component amount D1 based on the relationship between the signal waveform due to the periodic fluctuation component related to flicker degradation and the waveform of the filter coefficient of the bandpass filter that extracts the periodic fluctuation component. A band-pass vector phase correlation degree D34 representing the magnitude of a periodic time fluctuation component related to flicker degradation with respect to the time fluctuation component is detected.

  The fluctuation period component occupancy calculator 78 processes the bandpass vector phase correlation degree D34 to detect the fluctuation period component occupancy D10. That is, as shown in FIG. 25, in the fluctuation period component occupancy calculation unit 78, the vector phase correlation degree delay signal generation unit 79 sequentially delays the bandpass vector phase correlation degree D34 and outputs it by a plurality of systems. The maximum value detector 80 receives a plurality of output signals D35 and bandpass vector phase correlation D34 output from the vector phase correlation delay signal generator 79, and detects the maximum value from these input values. The fluctuation period component occupancy D10 is output. Also, the timing at which the maximum value is detected from the bandpass vector phase correlation degree D34 is notified to the synchronization bandpass component generation unit 81 (FIG. 21) by the maximum value detection synchronization signal D21.

  The synchronization bandpass component generation unit 81 acquires the normalized bandpass dot product signal D32 at the timing when the maximum value is detected from the bandpass vector phase correlation degree D34 by the maximum value detection synchronization signal D21, and the fluctuation feature amount generation unit 82. As a result, the synchronous bandpass component generation unit 81 detects and outputs the energy value of the time variation component related to flicker degradation.

  The variation feature amount generation unit 82 is configured in the same manner as the variation feature amount generation unit 23 of the first embodiment except that it operates based on the output signal D36 of the synchronous bandpass component generation unit 81 instead of the energy value D7. The

  According to the above configuration, the relationship between the signal waveform due to the periodic fluctuation component related to flicker degradation and the waveform of the filter coefficient of the band pass filter that extracts the periodic fluctuation component, the total time fluctuation component of the high-frequency component amount, Even if the bandpass vector phase correlation degree indicating the magnitude of the periodic time fluctuation component related to flicker deterioration is detected, and the highband component fluctuation feature quantity is obtained using this bandpass vector phase correlation degree, The same effect as the embodiment can be obtained.

  FIG. 26 is a block diagram illustrating a configuration of a high-frequency component variation feature amount measurement unit applied to the video signal processing device according to the fourth embodiment of the present invention. The video signal processing apparatus according to the fourth embodiment is configured in the same manner as the video signal processing apparatuses according to the respective embodiments described above except that the configuration of the high-frequency component variation feature quantity measuring unit 94 is different.

  Here, in the high frequency component variation feature amount measuring unit 94, the high frequency component amount delay unit 95 delays and outputs the high frequency component amount D1 for a period of at least one frame. The output signal D40 of the band component amount delay unit 95 is subtracted from the high band component amount D1 and output. The variation type signal generation unit 97 sets a reference value based on the output signal D40 of the high frequency component amount delay unit 95, determines the output signal D41 of the subtraction circuit 96, and indicates a variation that is a determination result of increase or decrease of the high frequency component amount D1. A type signal D42 is output.

  That is, FIG. 27 is a block diagram showing the variation type signal generation unit 97. In the variation type signal generation unit 97, the threshold generation unit 99 multiplies the output signal D40 of the high frequency component amount delay unit 95 by a positive coefficient smaller than 1 to generate a threshold value Th1 for determination.

  The variation type setting unit 98 determines the output signal D41 of the subtraction circuit 96 based on the threshold Th1, and generates a variation type signal D42. Specifically, when the value of the output signal D41 is DiffNL, the value of the threshold Th1 is EpsTh, and when DiffNL <−EpsTh, it is determined that the value is decreasing. Further, when DiffNL> EpsTh, it is determined that there is an increase. When EpsTh ≧ DiffNL ≧ −EpsTh, it is determined that there is no change. The variation type setting unit 98 outputs the determination result of decrease, increase and no variation by the variation type signal D42.

  The increasing fluctuation amount calculation unit 101 (FIG. 26) indicates that the fluctuation type signal D42 increases during a certain period, which is a period of a plurality of sampling periods of the output signal D41, in the direction reverse to the time axis from the present time. The sum of absolute values of the output signal D41 is calculated and output as an increasing fluctuation amount D43. This fixed period is hereinafter referred to as a processing window. As a result, as shown in FIGS. 28A and 28B, when the processing window is set during the period of 6 sampling periods of the output signal D41 and the original time point is the time point t, the fluctuation amount at the time of increase is calculated. The unit 101 performs sampling values D [t], D [t-1], and D of the output signal D41 input at time points t, t-1, t-2, t-3, t-4, and t-5, respectively. From [t-2], D [t-3], D [t-4], D [t-5], sampling values D [t] and D [t-3] at time points t and t-3 And the sum of the absolute values of the sampling values D [t] and D [t-3] is output as an increasing fluctuation amount D43.

  On the other hand, the fluctuation amount calculation unit 102 at the time of decrease indicates that the fluctuation type signal D42 has decreased from a plurality of sampling values of the output signal D41 included in the processing window, corresponding to the fluctuation amount calculation unit 101 at the time of increase. The sum of absolute values of the output signal D41 is calculated and output as a decrease amount of variation D44. Therefore, in the example of FIG. 28B, the sum of the absolute values of the sampling values D [t-2] and D [t-5] at time points t-2 and t-5 is output as the decreasing fluctuation amount D44. Instead of the absolute value sum, the increase variation D43 and the decrease variation D44 may be obtained by the square sum of the square sum, the square sum, or the like.

  The fluctuation cycle pattern detection unit 103 outputs a fluctuation cycle pattern detection flag F3 indicating the presence / absence of a periodic fluctuation of a high frequency component related to flicker deterioration based on the fluctuation type signal D42, the increase fluctuation amount D43, and the decrease fluctuation amount D44. .

  That is, FIG. 29 is a block diagram showing the fluctuation cycle pattern detection unit 103. In the fluctuation cycle pattern detection unit 103, the pattern perfect match judgment unit 105 determines the judgment result of the fluctuation type signal D42 included in the processing window and the judgment result based on the periodic time fluctuation pattern assumed in the high frequency component amount D1. Through this logical operation processing, it is determined whether or not the variation pattern specified by the variation type signal D42 is completely coincident with the assumed periodic time variation pattern. That is, for example, assuming that the assumed periodic time fluctuation pattern is (decrease, no change, increase, decrease, no change, increase), the time points t, t−1, t−2, t− in FIG. 3, t-4, t-5 determination results R [t], R [t-1], R [t-2], R [t-3], R [t-4], R [t-5 ] Completely matches the expected pattern of periodic time fluctuations. On the other hand, determination results R [t + 1], R [t], R [t-1] at time points t + 1, t, t-1, t-2, t-3, and t-4 obtained at subsequent sampling points. Since R [t-2], R [t-3], and R [t-4] are (no change, increase, decrease, no change, increase, decrease), in this case, it is determined that they are inconsistent. . Note that the assumed periodic time variation pattern is not limited to this example.

  The variation pattern symmetry determination unit 106 determines the symmetry of the sampling values determined to increase and decrease included in the processing window, and outputs a variation pattern symmetry determination flag F2. That is, the variation pattern symmetry determination unit 106 obtains the absolute value of the difference value between the increasing variation amount D43 and the decreasing variation amount D44. Further, the sum of the increasing fluctuation amount D43 and the decreasing fluctuation amount D44 is obtained, and a predetermined threshold value is obtained from the value of this sum. The variation pattern symmetry determination unit 106 determines the absolute value of the difference value based on the threshold value, and sets the variation pattern symmetry determination flag F2 based on the determination result.

  Specifically, when the increasing fluctuation amount D43 is INC, the decreasing fluctuation amount D44 is DEC, and the threshold is Vth, the threshold Vth is calculated by Vth = (INC + DEC) × Kv. Here, Kv is a positive coefficient less than 1. If | INC-DEC | <Vth, it is determined that there is symmetry, and the variation pattern symmetry determination flag F2 is set. In other cases, it is determined that there is no symmetry, and the variation pattern symmetry determination flag F2 is set. As a result, the variation pattern symmetry determination unit 106 determines symmetry when the increase variation amount D43 and the decrease variation amount D44 indicate substantially the same variation amount. Note that the method for determining symmetry is not limited to this, and the ratio of the variation amount D43 during increase and the variation amount D44 during decrease may be used.

  The degree-of-match determination unit 107 performs a logical operation process on the pattern complete match flag signal F1 and the variation pattern symmetry determination flag F2, and the pattern complete match flag signal F1 output from the pattern complete match determination unit 105 is a perfect match. When the fluctuation pattern symmetry determination flag F2 is symmetric, it is determined that a periodic fluctuation of a high frequency component related to flicker degradation, which is an assumed periodic temporal fluctuation pattern, has occurred, and the fluctuation period The pattern detection flag signal F3 is set with detection. If this determination result cannot be obtained, the fluctuation cycle pattern detection flag signal F3 is set without detection.

  The fluctuation feature quantity generation unit 108 (FIG. 26) processes the increase fluctuation quantity D43 and the decrease fluctuation quantity D44 with reference to the fluctuation cycle pattern detection flag F3 to generate a fluctuation feature quantity D12. Here, FIG. 30 is a block diagram showing the fluctuation feature quantity generation unit 108. In the variation feature amount generation unit 108, the variation amount calculation unit 109 adds the variation amount D43 during increase and the variation amount D44 during decrease to add a high frequency component when the variation period pattern detection flag F3 is set to be detected. The fluctuation amount D45 is output. On the other hand, when the fluctuation cycle pattern detection flag F3 is set without detection, the high frequency component fluctuation amount D45 is set to a value of 0 and output.

  The fluctuation amount delay unit 110 sequentially delays the high-frequency component fluctuation amount D45 by a predetermined delay time feature amount and outputs it by a plurality of systems. In this case, the plurality of systems may be one system. The delay time is longer than the length of the processing window described above.

  The maximum value detection unit 111 detects the maximum value from the output signal D46 of the variation amount delay unit 110 and the high frequency component variation amount D45 output from the variation amount calculation unit 109, and outputs the maximum value as the variation feature amount D12.

  In this embodiment, the increase / decrease of the high-frequency component amount D1 is determined, and based on the determination result, the match / mismatch of the increase / decrease pattern of the high-frequency component amount D1 with respect to the temporal variation pattern related to flicker deterioration is determined. When the determination result is obtained, the high-frequency component amount D1 is selectively processed to detect the high-frequency component variation feature amount D2, thereby calculating the high-frequency component variation feature amount D2 by simple processing. The same effect as the embodiment can be obtained.

  In addition, in the processing of the selective high frequency component amount D1 when this coincidence determination result is obtained, the increasing fluctuation amount and the decreasing fluctuation amount, which are the addition values of the high frequency component amount at the time of increase and decrease, respectively. By detecting and determining the symmetry of increase / decrease in the amount of high frequency component, the time variation of the high frequency component can be reduced when the occurrence of flicker degradation is reliably predicted.

  FIG. 31 is a block diagram illustrating a configuration of a high-frequency component variation feature amount measurement unit applied to the video signal processing device according to the fifth embodiment of the present invention. The video signal processing apparatus according to the fifth embodiment is configured in the same manner as the video signal processing apparatus 1 according to the first embodiment except that the configuration of the high-frequency component variation feature amount measuring unit 124 is different. In the high frequency component variation feature quantity measuring unit 124, the same components as those in the above-described embodiments are denoted by the corresponding reference numerals, and redundant description is omitted.

  In the high frequency component variation feature value measuring unit 124, the high frequency component amount delay unit 46 sequentially delays the high frequency component amount D1 and outputs the delayed high frequency component amount D1 through the 3n-1 system. Accordingly, the high-frequency component variation feature amount measuring unit 124 uses the continuous 3n sampling period based on the output signal D23 and the high-frequency component amount D1 of the high-frequency component amount delay unit 46 as a processing window, and thereby uses the high-frequency component amount D1. Process. Here, n is an integer, and in the processing in the high-frequency component variation feature quantity measurement unit 124, when securing the characteristics of FIGS. 4, 5, and 6, 3n is a value 3, a value 6, and a value 5 respectively. Set to In the following, a case where 3n is set to a value 6 and six sampling periods are set as processing windows will be described.

  The separation level setting unit 126 receives the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1, and detects a separation level D53 that is a predetermined processing reference value. Here, FIG. 32 is a block diagram showing a configuration of the separation level setting unit 126. In the separation level setting unit 126, the sorting processing unit 127 is a so-called order statistical filter, and the 3n sampling values based on the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1 are ordered in descending order. Or sort in ascending order of value. Here, as shown in FIG. 33, at the current time point t, the sorting processing unit 127 receives the high frequency component amount D1 at time points t, t-1, t-2, t-3, t-4, and t-5. Sampling values of certain values R [t], R [t-1], R [t-2], R [t-3], R [t-4], R [t-5] are input. Accordingly, in this case, the sorting processing unit 127 determines that these values R [t], R [t-1], R [t-2], R [t-3], R [t-4], R [t- 5] is sampled.

  Further, the sorting processing unit 127 outputs the second and third sampling values from the sorting result as the large intermediate level signal D51 and the small intermediate level signal D52, respectively, from the larger value side. Here, in the assignment of the sorting result to the large intermediate level signal D51 and the small intermediate level signal D52, the periodic time variation of the high frequency component amount D1 corresponds to the large intermediate level signal D51 and the small intermediate level signal D52. This is based on the assumption that the fluctuation is across two high-frequency component amounts, large and small. Therefore, in the example of FIG. 33, values R [t-5] and R [t-3], which are sampling values at time points t-5 and t-3, are output as the large intermediate level signal D51 and the small intermediate level signal D52, respectively. The

  The separation level calculation unit 128 calculates an average value from the large intermediate level signal D51 and the small intermediate level signal D52, and outputs this average value as the separation level D53. Therefore, in the example of FIG. 33, the separation level D53 (SPL [t]) is (R [t-5] + R [t-3]) / 2.

  The large high frequency component analyzing unit 130 (FIG. 31) sets a predetermined threshold value based on the separation level D53, and analyzes the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1 based on the threshold value. A large high frequency component fluctuation amount D55 indicating a fluctuation amount when the high frequency component amount D1 is large is detected.

  That is, FIG. 34 is a block diagram showing the large high frequency component analysis unit 130. In the large and high frequency component analyzing unit 130, the threshold setting unit 131 multiplies the separation level D53 by a coefficient of 1 or more to calculate a large and high frequency component determination threshold Th2 (LNTH [t] (see FIG. 33)).

  The large and high frequency component determination unit 132 sequentially determines 3n sampling values based on the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1 based on the large and high frequency component determination threshold Th2, and the determination result Is output by the large and high frequency component determination flag F4. In the following, a sampling value larger than the large high frequency component determination threshold Th2 is referred to as a large high frequency component.

  The counter unit 133 counts the number of large and high frequency components included in one processing window by counting the large and high frequency component determination flag F4 for each processing window, and calculates the count value as the number of large and high frequency component data. Output as D54.

  The large and high frequency component fluctuation amount calculating unit 134 performs sampling in which the large and high frequency component determination flag F4 indicates the large and high frequency component with respect to the output signal D23 and the high frequency component amount D1 of the high frequency component amount delay unit 46. The sum of absolute differences with respect to the separation level D53 is obtained by the value, the sum of absolute differences is divided by the count value of the counter unit 133, and the resulting divided value is output as the large and high frequency component fluctuation amount D55. That is, in the example of FIG. 33, at time t, sampling values R [t], R [t-1], R [at time t, t-1, t-2, t-3, t-4, t-5. t-2], R [t-3], R [t-4], and R [t-5], the separation level D53 of the value SPL [t] is obtained and generated based on this separation level D53. By determining these sampling values R [t], R [t-1], R [t-2], R [t-3], R [t-4], R [t-5] based on the threshold value Th2, Sampling values R [t-2] and R [t-5] of t-2 and t-5 are determined to be large and high frequency components. Accordingly, in this case, the large and high frequency component fluctuation amount calculation unit 134 calculates the sum of absolute differences of the sampling values R [t−2] and R [t−5] from the separation level D53 by the arithmetic processing of the following equation. By dividing by the count value (value 2) of the counter unit 133, the large and high frequency component fluctuation amount D55 is generated by the average value INV [t] of the absolute value of the difference from the separation level D53.

  Similarly to the large and high frequency component analyzing unit 130, the small and high frequency component analyzing unit 135 sets a predetermined threshold value by the separation level D53, and the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount are set based on this threshold value. D1 is analyzed, and a small high frequency component fluctuation amount D57 indicating a fluctuation amount when the high frequency component amount D1 is small is detected.

  That is, FIG. 35 is a block diagram showing the small high-frequency component analysis unit 135. In the small and high frequency component analysis unit 135, the threshold setting unit 136 multiplies the separation level D53 by a coefficient of value 1 or less to calculate the small and high frequency component determination threshold Th3 (SNTH [t] (see FIG. 33)).

  The small and high frequency component determination unit 137 sequentially determines 3n sampling values based on the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1 based on the small and high frequency component determination threshold Th3, and determines the determination result. Output by the region component determination flag F5. Hereinafter, a sampling value smaller than the small high frequency component determination threshold Th3 is referred to as a small high frequency component.

  The counter unit 138 counts the number of small high frequency components included in one processing window by counting the small high frequency component determination flag F5 for each processing window, and outputs the count value as the small high frequency component data number D56. To do.

  The small high frequency component fluctuation amount calculation unit 139 separates the output signal D23 of the high frequency component amount delay unit 46 and the high frequency component amount D1 by the sampling value in which the small high frequency component determination flag F5 indicates the small high frequency component. The sum of absolute differences from the level D53 is obtained, the sum of absolute differences is divided by the count value of the counter unit 138, and the average value of the sum of absolute values obtained as a result is output as the small and high range component fluctuation amount D57. That is, in the example of FIG. 33, at time t, sampling values R [t], R [t-1], R [at time t, t-1, t-2, t-3, t-4, t-5. t-2], R [t-3], R [t-4], R [t-5], and a threshold value Th3 (SNTH [t]) is generated from the separation level D53 (SPL [t]). The sampling values R [t], R [t-1], R [t-2], R [t-3], R [t-4], and R [t-5] based on the threshold Th3 are as follows: Sampling values R [t], R [t-1], R [t-3], and R [t-4] at time points t, t-1, t-3, and t-4 are determined as small high-frequency components. . Accordingly, in this case, the small and high frequency component fluctuation amount calculation unit 139, as shown by the following equation, the sampling values R [t], R [t-1], R [t-3], R [t-4] The difference absolute value sum with the separation level D53 is divided by the count value (value 4) of the counter unit 138, and the small high frequency component fluctuation amount D57 is output by the average value SNV [t] of the difference absolute values. In place of the average value calculation processing based on the sum of absolute differences in the large and high frequency component fluctuation amount calculating unit 134 and the small and high frequency component fluctuation amount calculating unit 139, the average value is calculated by the mean square difference, the square root of the mean square difference, or the like. You may make it ask.

  The variation feature reliability setting unit 140 (FIG. 31) processes the large and high frequency component data number D54 and the small and high frequency component data number D56 to detect the variation feature reliability D60. FIG. 36 is a block diagram showing the variation feature reliability setting unit 140. In this variation feature reliability setting unit 140, the variation determination unit 141 has a small time variation of the high frequency component amount D1 when either the large high frequency component data number D54 or the small high frequency component data number D56 is 0. The fluctuation determination flag F6 is set without fluctuation indicating that. In other cases, the variation determination flag F6 is set with variation indicating that the temporal variation of the high frequency component amount D1 is large.

  The high frequency component composition ratio calculation unit 142 calculates the ratio between the small high frequency component data number D56 and the large high frequency component data number D54 when the fluctuation determination flag F6 is set to have a fluctuation, and calculates the calculated ratio. Output as a high frequency component composition ratio D58.

  The variation feature reliability calculation unit 143 outputs the variation feature reliability D59 with a value of 0 when the variation determination flag F6 is set to have no variation. On the other hand, when the variation determination flag F6 is set to have variation, a value is set according to the high frequency component composition ratio D58 and the variation feature reliability D59 is output according to the characteristics shown in FIG. That is, the fluctuation feature reliability calculation unit 143 determines the high frequency component composition ratio D58 by the threshold values RNTh1 and RNTh2, and when the high frequency component composition ratio D58 is within the range determined by the threshold values RNTh1 and RNTh2, the fluctuation feature reliability D59 is obtained. Set to the value 1. When the high-frequency component composition ratio D58 is outside the range determined by the threshold values RNTh1 and RNTh2, the variation feature reliability D59 is set so as to gradually approach the value 0 as the value increases away from the threshold values RNTh1 and RNTh2. Here, the threshold values RNTh1 and RNTh2 are set according to the nature of the periodic time variation of the assumed high frequency component amount D1.

  The reliability time smoothing unit 144 smoothes the variation feature reliability D59 and outputs a variation feature reliability D60. Note that an FIR filter, IIR filter, median filter, or the like can be applied to the reliability time smoothing unit 144.

  When the variation feature reliability D60 is equal to or less than a predetermined threshold, the variation feature amount generation unit 145 determines that the periodic feature variation of the high frequency component amount D1 related to flicker degradation has not occurred and sets the variation feature amount D12. Set to 0. In other cases, the sum of the large high frequency component variation amount D55 and the small high frequency component variation amount D57 is set as the variation feature amount D12 and output.

  According to the above configuration, the high frequency component amount obtained in a certain period is processed by the order statistical filter, the separation level used for the determination of the high frequency component amount is set, and the high frequency component is determined by this separation level. Even if the high-frequency component variation feature value is generated, the same effect as in the first embodiment can be obtained.

  That is, in the process of generating the high-frequency component variation feature value, the high-frequency component amount is determined based on the separation level, the average value of the difference from the separation level of the high-frequency component amount having a large value, and the high-frequency component having a small value In addition to detecting the average value of the difference from the separation level of the quantity, the ratio of the number of samplings used to calculate these average values is detected, and the average value due to the high-frequency component amount having a large value is detected as the high-frequency component having a small value. Even if a signal level change similar to a picture type periodic fluctuation component occurs temporarily by correcting the average value of Even when the high frequency component of the video signal changes variously, flicker degradation can be reliably suppressed.

  FIG. 38 is a block diagram showing a video signal processing apparatus according to Embodiment 6 of the present invention. In the video signal processing apparatus 151 of the sixth embodiment, the same components as those of the video signal processing apparatus 1 described above with reference to FIG.

  Here, the video signal processing device 151 processes the high frequency component amount D1 output from the high frequency component measuring unit 3 simultaneously and in parallel by a plurality of high frequency component variation feature amount measuring units 154A to 154N. Component variation feature amounts D2A to D2N are detected. Here, these high-frequency component variation feature amount measurement units 154A to 154N are the same as the high-frequency component variation feature amount measurement units of Examples 1 to 5 described above, except that the periodic variation components of the high-frequency components to be detected are different. Configured identically. As a result, in this video signal processing apparatus 151, even when various input video signals S1 having different GOP structures are processed, any one of the high frequency component variation feature amount measurement units 154A to 154N correctly performs the high frequency component variation feature. It is comprised so that quantity D2A-D2N can be detected.

  The fluctuation feature quantity selection unit 155 detects the maximum value from the high band component fluctuation feature quantities D2A to D2N output from the high band component fluctuation feature quantity measurement units 154A to 154N, and the high band component fluctuation feature quantity based on the maximum value. D2 is output to the video signal processing unit 2.

  According to this embodiment, a plurality of high-frequency component fluctuation feature quantity measurement units having different cyclic fluctuation components of the high-frequency component to be detected are detected simultaneously in parallel, and these high-frequency component fluctuations are detected. By detecting the maximum value from the feature quantity and controlling the video signal processing unit, it is possible to sufficiently reduce flicker degradation when processing various input video signals 1 having different GOP structures.

  FIG. 39 is a block diagram showing a video signal processing apparatus according to Embodiment 7 of the present invention. In this video signal processing device 161, the same configurations as those of the video signal processing devices of the above-described embodiments are denoted by the corresponding reference numerals, and redundant description is omitted. In the following description, the configuration of the video signal processing apparatus according to the seventh embodiment is shown on the premise of the configuration of the first embodiment. The video signal processing apparatus according to the seventh embodiment is different from the first embodiment. The present invention can also be applied when a configuration is assumed.

  In this video signal processing device 161, the partial area dividing unit 162 receives the input video signal S1, and sets one screen by the input video signal S1 by dividing it equally in the vertical and horizontal directions as shown in FIG. The input video signal S1 is divided and output for each area AR. Here, FIG. 40 shows an example in which the area AR is formed by dividing one screen into five equal parts in the horizontal direction and the vertical direction.

  The brightness classifying unit 163 aggregates the luminance levels of the video signal S7 output from the partial region dividing unit 162 for each region AR, and detects a partial region brightness level D71 for classifying the brightness of the region AR. Note that, for example, an average value and a total value of luminance levels are applied to the aggregation. Thereby, the brightness classification unit 163 classifies each area AR of the input video signal S1 by brightness.

  The texture classification unit 164 receives the luminance level of the video signal S7 output from the partial region dividing unit 162, and classifies the region AR by texture. Specifically, as shown in FIG. 40, the texture classification unit 164 outputs from the partial region dividing unit 162 for each sub-region ARA set by equally dividing each region AR in the vertical direction and the horizontal direction. The variance SubRVer of the video signal S7 is calculated. Further, this variance SubRVer is averaged for each area AR to obtain an average AveSubRVar of the variance SubRVer. Also, the variance VarSubRVar of the variance SubRVer is calculated for each area AR.

  The texture classification unit 164 determines the average AveSubRVar and variance VarSubRVar for each area AR, and classifies the texture of each area AR. That is, when the variance VarSubRVar is larger than the predetermined threshold TxrTh, the texture classification unit 164 distributes the variance SubRVer in the sub-region ARA unevenly in the region AR, and the region AR has different textures when there is an edge. Therefore, the partial texture classification information D72 of the area AR is set to non-texture.

  Also, when the variance VarSubRVar is smaller than the predetermined threshold TxrTh, as the average AveSubRVar value increases, the region is sequentially classified into noise type, small amplitude texture, medium amplitude texture, and large amplitude texture, and the partial region texture classification information D72 is set.

  Note that the classification method is not limited to this, and a frequency analysis or the like may be performed to extract more feature amounts in the partial input video signal S7, and further classification may be performed, or noise, texture, etc. Alternatively, it may be configured to be simply classified in two stages. Thereby, the texture classification unit 164 classifies each area AR of the input video signal S1 by a classification method using a variation in pixel values different from that of the brightness classification unit 163 as a determination criterion.

  Accordingly, the video signal processing apparatus classifies the area AR set in the input video signal S1 according to the feature amount of the input video signal S1.

  The high-frequency component measurement unit 165 and the high-frequency component variation feature amount measurement unit 166 perform the high-frequency component amount D1 and the high-frequency component variation feature amount for each classification classified by the partial region brightness level D71 and the partial texture classification information D72. D2 is detected. Therefore, when the entire area AR related to the input video signal S1 is classified into one classification, the high frequency component amount D1 and the high frequency component variation feature amount D2 are detected on one screen as in the above-described embodiments. . The high-frequency component measurement unit 165 and the high-frequency component variation feature amount measurement unit 166 are different from the above-described embodiments except that the configurations regarding the detection units of the high-frequency component amount D1 and the high-frequency component variation feature amount D2 are different. Configured identically.

  As shown in FIG. 41, the video signal processing unit 167 receives control information D16, D17, and D18 for controlling the noise reduction processing unit 32, the sharpness enhancement processing unit 33, and the contrast enhancement processing unit 34 in the control information generation unit 168. These processing units 32 to 34 are generated so as to reduce the flicker deterioration generated by the high frequency component amount D1 and the high frequency component variation feature amount D2.

  Here, the control information generation unit 168 generates control information D16, D17, and D18 for each classification classified by the partial region brightness level D71 and the partial texture classification information D72. At this time, the partial region brightness level D71 in consideration of the human visibility characteristics with respect to the area, spatial frequency, and brightness of each classification, the visibility characteristics with respect to the temporal variation to be detected in the high-frequency component variation feature amount D2, and the like. The settings of the control information D16, D17, and D18 in each classification may be changed by the partial texture classification information D72 and the like.

  Specifically, since the flicker deterioration is more noticeable in the dark portion, the control information D16, D17 is set so that the dark portion with the partial region brightness level D71 increases the flicker deterioration reduction effect with respect to the high-frequency component variation feature amount D2. , D18 can be set. Further, since the flicker deterioration is less noticeable in the small area portion than in the large area portion, the control information D16 is set so that the flicker deterioration reduction effect with respect to the high frequency component variation feature amount D2 is low in the small area portion. , D17, and D18 can be set. In addition, when a high frequency component related to flicker degradation is included, flicker degradation is more conspicuous than when this type of high frequency component is small. Therefore, the effect of reducing flicker degradation as the spatial frequency increases. It is conceivable to set the control information D16, D17, D18 so as to be high.

  The partial area integration unit 169 (FIG. 39) rearranges the video signals S2 thus processed for each area AR in the order of the original input video signals S1 and outputs them.

  In this embodiment, each region set in the input video signal is classified by various feature amounts, and by setting the high frequency component amount D1 and the high frequency component variation feature amount D2 for each classification, the flicker degradation is more reliably performed. Can be reduced. In addition, the image quality of the output video signal can be improved.

  FIG. 42 is a block diagram showing a video signal processing apparatus according to the eighth embodiment of the present invention. This video signal processing device 171 measures the magnitude of subjective degradation of flicker degradation perceived on the display. The video signal processing device 171 is configured in the same manner as each of the embodiments described above except that a flicker deterioration subjective perception level setting unit 172 is provided instead of the video signal processing unit 167. In the configuration of the seventh embodiment, a configuration in which a flicker deterioration subjective perception level setting unit 172 is provided instead of the video signal processing unit 167 will be described.

  The flicker degradation subjective perception level setting unit 172 performs the subjective perception of flicker degradation in the same manner as when calculating the control amount based on the control information related to the contrast enhancement processing in the control information generation of the video signal processing unit of each embodiment described above. The degree is measured and output as a measurement result D81. Here, the subjective perception level of flicker degradation is an index indicating the degree of flicker degradation that occurs when displayed on a display, and depends on human visibility characteristics and the like. Accordingly, in this embodiment, the flicker degradation subjective perception level setting unit 172 is similar to the control information generation unit 168 described above with respect to the seventh embodiment, and includes the high frequency component amount D1, the high frequency component variation feature amount D2, and the partial region brightness. A measurement result D81 is output based on the level D71 and the partial texture classification information D72.

  Further, the flicker degradation subjective perception level setting unit 172 receives the non-linear characteristic information D80 until the input video signal S1 is displayed on the display monitor, and corrects and outputs the measurement result based on this information. Here, the non-linear characteristic information D80 includes characteristic information such as non-linear processing performed by the photographing apparatus, non-linear processing for contrast enhancement, and non-linear processing considering characteristics of the display monitor.

  According to the above configuration, the amount of high frequency component contained in the input video signal is periodically detected, and the subjective perception of flicker deterioration is measured based on the detection result, thereby quantifying flicker deterioration. Can be measured automatically.

  In the above-described embodiment, the case where the video signal processing apparatus is configured by the processor has been described. However, the present invention is not limited to this, and can be widely applied to the case where the video signal processing apparatus is configured by a hardware configuration. .

  The present invention can be applied to processing of a video signal obtained by decoding streaming data such as MPEG.

It is a block diagram which shows the high frequency component fluctuation | variation feature-value measurement part of the video signal processing apparatus of Example 1 of this invention. It is a block diagram which shows the video signal processing apparatus of Example 1 of this invention. It is a characteristic curve figure with which it uses for description of the GOP structure of input video signal S1. It is an approximate line figure used for explanation of a band pass filter part. FIG. 5 is a schematic diagram for explaining an example different from FIG. 4. FIG. 6 is a schematic diagram for explaining an example different from FIGS. 4 and 5. It is a characteristic curve figure with which it uses for description of the relationship between a band pass filter part and a high pass filter part. It is a block diagram which shows the structure of a high pass filter part. It is a characteristic curve figure which shows the relationship between a fluctuation period component occupation degree and a fluctuation period stability. It is a block diagram which shows a fluctuation | variation feature-value production | generation part. It is a characteristic curve figure which shows occupancy reliability. It is a characteristic curve figure which shows stability reliability. It is a block diagram which shows a video signal processing part. It is a characteristic curve figure with which it uses for description of control of a contrast emphasis processing part. It is a block diagram which shows the structure of the high region component fluctuation | variation feature-value measurement part applied to the video signal processing apparatus of Example 2 of this invention. It is a block diagram which shows a band pass filter output energy calculation part. It is a characteristic curve figure which shows the base waveform of a band pass filter. It is a characteristic curve figure with which it uses for description of a high region component amount. It is a block diagram which shows a high region component energy calculation part. It is a basic diagram which shows each component in an orthonormal coordinate space. It is a block diagram which shows the high region component fluctuation | variation feature-value measurement part applied to the video signal processing apparatus of Example 3 of this invention. It is a block diagram which shows an AC component vector component part. It is a block diagram which shows a band pass vector phase correlation calculation part. It is a basic diagram which shows each component on an orthonormal coordinate space. It is a block diagram which shows a fluctuation period component occupation degree calculation part. It is a block diagram which shows the high frequency component fluctuation | variation feature-value measurement part applied to the video signal processing apparatus of Example 4 of this invention. It is a block diagram which shows a fluctuation type signal production | generation part. It is a characteristic curve figure with which it uses for description of determination of increase / decrease in a high region component amount. It is a block diagram which shows a fluctuation period pattern detection part. It is a block diagram which shows a fluctuation | variation feature-value production | generation part. It is a block diagram which shows the high frequency component fluctuation | variation feature-value measurement part applied to the video signal processing apparatus of Example 5 of this invention. It is a block diagram which shows a separation level setting part. It is a characteristic curve figure with which it uses for description of operation | movement of a separation level setting part. It is a block diagram which shows a large high region component analysis part. It is a block diagram which shows a small high region component analysis part. It is a block diagram which shows a fluctuation characteristic reliability setting part. It is a characteristic curve figure with which it uses for description of operation | movement of a fluctuation characteristic reliability setting part. It is a block diagram which shows the video signal processing apparatus of Example 6 of this invention. It is a block diagram which shows the video signal processing apparatus of Example 7 of this invention. It is a basic diagram with which it uses for description of the division | segmentation of 1 screen of an input video signal. It is a block diagram which shows a video signal processing part. It is a block diagram which shows the video signal processing apparatus of Example 8 of this invention.

Explanation of symbols

1, 151, 161, 171... Video signal processing device, 2, 167... Video signal processing unit, 3, 165... High frequency component measuring unit, 4, 44, 64, 94, 124, 154A to 154N, 166 ...... High-frequency component variation feature quantity measurement unit, 11... Band-pass filter unit, 12... High-pass filter unit, 18, 48... Band-pass output energy calculation unit, 19. , 78... Fluctuation period component occupancy calculation unit, 21... Variation period stability calculation unit, 23, 82, 108, 145... Variation feature amount generation unit, 30. ... Control information generation unit, 49 ... Temporal AC energy calculation unit, 65 ... AC component vector component unit, 73 ... Bandpass vector phase correlation calculation unit, 81 ... Synchronous bandpass component generation unit, 97 ... Fluctuation type signal generation unit 101... Fluctuation amount calculation unit at increase 102. Fluctuation amount calculation unit at decrease 103... Variation period pattern detection unit 126. Analysis unit 135. Small and high frequency component analysis unit 140 140 Fluctuation feature reliability setting unit 155. Fluctuation feature amount selection unit 163... Brightness classification unit 164. Area combining unit, 172... Flicker deterioration subjective perception setting unit

Claims (10)

A video signal processing unit for processing an input video signal to generate an output video signal;
A high frequency component measuring unit for detecting a high frequency component amount included in the input video signal;
A high-frequency component variation feature amount measuring unit that detects a high-frequency component variation feature amount indicating a periodic temporal variation amount of the high-frequency component amount from the high-frequency component amount;
The video signal processor is
Based on the high-frequency component variation feature, the characteristic of the video signal processing is varied so as to suppress periodic variation in the high-frequency component of the output video signal ,
The high-frequency component variation feature amount measurement unit is
Bandpass output processing for extracting a signal component corresponding to a periodic pattern of the picture type set in the input video signal from the high frequency component amount;
A high-pass output process for extracting a high-frequency component including a signal component by the band-pass output process from the high-frequency component amount;
Correcting the signal component by the bandpass output process with the high-frequency component by the high-pass output process, and executing the high-frequency component variation feature value generating process for generating the high-frequency component variation feature value;
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The video signal processing unit increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement in the sharpness enhancement process Reduce nonlinearity in processing,
Video signal processing device. A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high frequency component variation characteristic amount, so as to suppress periodic variations in the high frequency component of said output video signal, have a characteristic variable step for varying the characteristics of the video signal processing,
The high-frequency component variation feature amount measuring step includes:
A bandpass output step for extracting a signal component corresponding to a picture type periodic pattern set in the input video signal from the high frequency component amount;
A high-pass output step for extracting a high-frequency component including a signal component by the band-pass output step from the high-frequency component amount;
A signal component generated by the band pass output step is corrected by a high frequency component generated by the high pass output step to generate the high frequency component variation feature value, and a high frequency component variation feature value generation step,
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The characteristic variable step increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high-frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement process in the sharpness enhancement process Reduce nonlinearity in
Video signal processing method. The high frequency component variation feature value generation step includes:
A bandpass output energy calculating step of calculating an energy value of the signal component by the bandpass output step;
A high-pass output energy calculating step for calculating an energy value of a high-frequency component by the high- pass output step;
A fluctuation period component occupancy calculation step for calculating a fluctuation period component occupancy indicating a ratio between the energy value obtained in the band pass output energy calculation step and the energy value obtained in the high pass output energy calculation step;
A fluctuation period stability calculating step for calculating a fluctuation period stability indicating stability in the time axis direction of the fluctuation period component occupancy; and
A variation feature generation step of detecting the high-frequency component variation feature amount by correcting the energy value obtained in the bandpass output energy calculation step with the variation period component occupancy and the variation period stability.
The video signal processing method according to claim 2 . A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high-frequency component variation feature amount, a characteristic variable step of varying the characteristic of the video signal processing so as to suppress periodic variation in the high-frequency component of the output video signal,
The video signal processing step includes
By varying the characteristics according to the input video signal and amplifying the input video signal by nonlinear characteristics, the output video signal is generated by enhancing the contrast of the input video signal,
Based on the high-frequency component fluctuation feature amount, the nonlinear characteristic is brought close to a linear characteristic to suppress the periodic fluctuation ,
The characteristic variable step reduces nonlinearity in the contrast enhancement executed by the video signal processing step as the time variation amount indicated by the high-frequency component variation feature amount increases.
Film image signal processing method. A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high-frequency component variation feature amount, a characteristic variable step of varying the characteristic of the video signal processing so as to suppress periodic variation in the high-frequency component of the output video signal,
The high-frequency component variation feature amount measuring step includes:
Due to the relationship between the signal waveform of the signal component related to the periodic time variation and the waveform of the filter coefficient of the bandpass filter required for extraction of the signal component, the periodic time with respect to the time variation component of the high frequency component amount A bandpass vector phase correlation degree detecting step for detecting a bandpass vector phase correlation degree representing a magnitude of a signal component related to fluctuation;
A synchronous bandpass component generation step of detecting an energy value of a signal component related to the periodic time variation;
A fluctuation period component occupancy degree calculating step for calculating a fluctuation period component occupancy degree indicating a ratio of an energy value of the signal component to the high frequency component by the band pass vector phase correlation degree;
A fluctuation period stability calculating step for calculating a fluctuation period stability indicating stability in the time axis direction of the fluctuation period component occupancy; and
The energy value obtained in synchronous bandpass component generating step, and corrected by the fluctuation cycle component occupancy and the fluctuation cycle stability possess a fluctuation feature amount generating step of detecting the high frequency component variation characteristic amounts,
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The characteristic variable step increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high-frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement process in the sharpness enhancement process Reduce nonlinearity in
Film image signal processing method. A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high-frequency component variation feature amount, a characteristic variable step of varying the characteristic of the video signal processing so as to suppress periodic variation in the high-frequency component of the output video signal,
The high-frequency component variation feature amount measuring step includes:
A variation type signal generation step of determining increase / decrease in the high frequency component amount and outputting a variation type signal;
Whether or not the time variation pattern by the variation type signal matches the periodic time variation pattern assumed by the high frequency component amount, which is the picture type periodic pattern set in the input video signal A match determination step for determining whether or not
Based on the determination result of the match determination step, when the match determination result is obtained in the match determination step, the variation feature value generation that processes the high frequency component amount and detects the high frequency component variation feature value and a step to Yes,
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The characteristic variable step increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high-frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement process in the sharpness enhancement process Reduce nonlinearity in
Film image signal processing method. The variable feature generation step includes
Based on the time variation pattern, in the variation type signal generation step, an increasing variation amount detecting step of detecting the increasing variation amount by adding the high frequency component amount determined to be increasing;
Based on the time variation pattern, in the variation type signal generation step, a decrease variation amount detection step of detecting the decrease variation amount by adding the high frequency component amount determined to be decreased; and
A symmetry determining step for determining symmetry of increase / decrease of the high-frequency component amount based on the increasing variation amount and the decreasing variation amount;
Based on the determination result in the coincidence determination step and the determination result in the symmetry determination step, the variation amount at the time of increase is set by adding the variation amount at increase and the variation amount at decrease. The video signal processing method according to claim 6 , further comprising a processing step. The variable feature generation step includes
A separation level setting step for sorting the high frequency component amounts obtained in a certain period, and setting a separation level used for the determination of the high frequency component amount according to the high frequency component amounts in a predetermined order of the sorting results;
The video signal processing method according to claim 7 , further comprising: a processing step based on a separation level for determining the high-frequency component amount based on the separation level and detecting the high-frequency component variation feature amount. A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high frequency component variation characteristic amount, so as to suppress periodic variations in the high frequency component of said output video signal, have a characteristic variable step for varying the characteristics of the video signal processing,
The high-frequency component variation feature amount measuring step includes:
A bandpass output step for extracting a signal component corresponding to a picture type periodic pattern set in the input video signal from the high frequency component amount;
A high-pass output step for extracting a high-frequency component including a signal component by the band-pass output step from the high-frequency component amount;
A signal component generated by the band pass output step is corrected by a high frequency component generated by the high pass output step to generate the high frequency component variation feature value, and a high frequency component variation feature value generation step,
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The characteristic variable step increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high-frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement process in the sharpness enhancement process Reduce nonlinearity in
Video signal processing method program. In a recording medium on which a program for a video signal processing method is recorded,
The program is
A video signal processing step of processing an input video signal to generate an output video signal;
A high frequency component measuring step for detecting a high frequency component amount included in the input video signal;
A high frequency component variation feature amount measuring step for detecting a high frequency component variation feature amount indicating a periodic temporal variation amount of the high frequency component amount from the high frequency component amount, and
The video signal processing step includes
Based on the high frequency component variation characteristic amount, so as to suppress periodic variations in the high frequency component of said output video signal, have a characteristic variable step for varying the characteristics of the video signal processing,
The high-frequency component variation feature amount measuring step includes:
A bandpass output step for extracting a signal component corresponding to a picture type periodic pattern set in the input video signal from the high frequency component amount;
A high-pass output step for extracting a high-frequency component including a signal component by the band-pass output step from the high-frequency component amount;
A signal component generated by the band pass output step is corrected by a high frequency component generated by the high pass output step to generate the high frequency component variation feature value, and a high frequency component variation feature value generation step,
The video signal processing includes noise reduction processing, sharpness enhancement processing or contrast enhancement processing,
The characteristic variable step increases the noise reduction effect in the noise reduction process as the time fluctuation amount indicated by the high-frequency component fluctuation feature quantity is larger, or the amplification factor or the contrast enhancement process in the sharpness enhancement process Reduce nonlinearity in
A recording medium on which a program for a video signal processing method is recorded.

Images