JP2006270417A - Video signal processing method and video signal processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a video signal processing method and a video signal processing apparatus for optimally correcting the image quality, according to the state of the video signal. <P>SOLUTION: The video signal processing method is configured, such that when correcting gradation of a video signal and its color signal component, a parameter used for at least correction of the gradation or the color signal component is supplied at a changed response speed, by changing the feedback coefficient value of a leak type integrating circuit 31, a characteristics extracting means 20 extracts the characteristics of the video signal by each field or each frame, a coefficient changing means 32 obtains variations that are the index of a scene change, on the basis of the extracted feature, and the feedback coefficient value of the leak integrating circuit is set by each field or each frame according to the variation; and when setting the feedback coefficient value, the feedback coefficient value is changed in the direction of increasing the response speed as the variation is greater within a prescribed range, wherein the variation is subject to change. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to video signal processing for improving visual image quality by correcting gradation and color signal components of a video signal, and in particular, automatically improves the image quality according to the characteristics of the image of the input video signal. The present invention relates to a video signal processing method and a video signal processing apparatus that can be adjusted.

  In television broadcasting and the like in which a display device that displays a video signal is assumed to be a cathode ray tube (CRT), a gamma characteristic is applied to the video signal in advance on the transmission side. In a display device using a cathode ray tube, the video signal is displayed in a linear gradation by combining the gamma characteristic provided on the transmission side with the inverse gamma characteristic of the cathode ray tube. On the other hand, the plasma display panel display device (PDP) and the liquid crystal display device (LCD) have a linear gradation characteristic, so that a video signal subjected to the gamma characteristic is displayed with a linear gradation. Requires gamma correction for applying an inverse gamma characteristic to the video signal.

  The gamma correction is used not only for the purpose of making the gradation linear, but also for the purpose of correcting the image quality by appropriately changing the gamma curve formed by the input luminance signal Y and the output luminance signal Y ′. Some gamma curves for correcting image quality have S-characteristics, and others add or subtract DC. Furthermore, there are some that add or subtract a direct current component to the S-shaped characteristic. As a device for performing such image quality correction, a gradation correction processing device is proposed in which the gain and gamma characteristics are not suppressed regardless of the dynamic range (for example, see Patent Document 1 below).

On the other hand, there are various types of television receivers using a plasma display panel display device or a liquid crystal display device, and the chromaticity points of the three primary colors vary even with the same model. Therefore, a color correction circuit that corrects hue, saturation, luminance, etc. for only a specific color and its neighboring colors is used. As the color correction circuit, the color difference signals RY and BY and the luminance signal Y are input, and the angle T of the color difference signals RY and BY is calculated first, and then the angle T is corrected using the angle T as a parameter. A saturation correction coefficient, a hue correction coefficient, a Y (luminance) gamma correction coefficient, and a C (saturation) gamma correction coefficient are calculated for each color according to the number of target colors, and then the saturation of each color is calculated. The degree correction coefficient, the hue correction coefficient, the Y gamma correction coefficient, the C gamma correction coefficient are added to each other, and the total saturation correction coefficient, the hue correction coefficient, Y gamma, The correction coefficient total value and the C gamma correction coefficient total value are obtained, and then the color difference signals RY and BY are calculated based on the saturation correction coefficient total value and the hue correction coefficient total value. A device for correcting saturation and hue is disclosed (for example, see Patent Document 2 below). ).
Japanese Patent Laid-Open No. 2002-27285 JP 2003-348614 A

  However, since the gamma curve in the above-described conventional gradation correction processing apparatus is a predetermined fixed one, there is a problem that the image quality of the image that changes momentarily cannot be optimized. In the above-described conventional color correction circuit, hue coefficients p1 and p2 for correcting the hue correction coefficient, Y gamma coefficient gy for correcting the Y gamma correction coefficient, and C gamma coefficient for correcting the C gamma correction coefficient. Since gc and the like are uniquely defined regardless of the video, there is a problem that in the case of an entirely dark scene with a poor S / N ratio, the noise is also corrected and the noise is conspicuous.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a video signal processing method and a video signal processing apparatus capable of optimally correcting the image quality according to the state of the video signal. It is to provide.
Another object of the present invention is to provide a video signal processing method and a video signal processing apparatus capable of correcting a color to be corrected vividly and preventing noise from standing out even in a dark scene. There is.

In order to achieve the above object, the present invention provides a video signal processing method for correcting gradation and color signal components of a video signal.
Supplying a parameter used for correcting at least one of the gradation and the color signal component by changing a response speed by changing a feedback coefficient value of a leaky integration circuit;
Extracting at least one of a histogram, an average luminance value, a minimum luminance value, a maximum luminance value, and a motion change amount as a feature of the video signal for each field or each frame;
Obtaining a variation value that is an index of a scene change based on the extracted feature, and setting a feedback coefficient value of the leaky integration circuit according to the variation value for each field or each frame,
The step of setting the feedback coefficient value is characterized by changing and setting the feedback coefficient value in a direction in which the response speed is increased as the fluctuation value increases within a predetermined range in which the fluctuation value changes.

The present invention also relates to a video signal processing apparatus comprising a gradation correction circuit for correcting the gradation of a video signal and a color correction circuit for correcting a color signal component.
A leak type integration circuit that inputs parameters used for correcting at least one of the gradation and the color signal component, and supplies the parameters while changing the response speed;
Feature extraction means for extracting at least one of a histogram, an average luminance value, a minimum luminance value, a maximum luminance value, and a motion change amount as a feature of the video signal for each field or each frame;
Based on the features extracted by the feature extraction means, a variation value that is an index of scene change is obtained, and a coefficient change that sets the feedback coefficient value of the leaky integration circuit for each field or frame according to the variation value Means and
The coefficient changing means changes and sets the feedback coefficient value in a direction in which the response speed increases as the fluctuation value increases within a predetermined range in which the fluctuation value changes.

  According to the present invention, even in a scene change in which a scene changes suddenly, the response speed can be increased by adaptively reducing the value of a parameter for deriving the feedback coefficient value of the leaky integration circuit. The image quality can be optimally corrected according to the signal state. In addition, by applying the same processing to the parameters used to correct the color signal components, the color to be corrected can be corrected vividly, and noise is not noticeable even in dark scenes. can do.

  Hereinafter, the present invention will be described in detail based on preferred embodiments shown in the drawings. FIG. 1 is a block diagram showing the configuration of an embodiment of a video signal processing apparatus for implementing a video signal processing method according to the present invention. In the figure, a luminance signal Y and color difference signals BY and RY, which are video signals, are input to the video processing unit 10. The video processing unit 10 includes a gradation correction circuit 11 that performs gamma correction on a video signal and a color correction circuit 12 that performs color correction. The luminance signal Y ′ and the color difference signal corrected by these circuits, respectively. (B−Y) ′ and (R−Y) ′ are output. In addition, the feature extraction unit 20 that detects the feature of the video signal input to the video processing unit 10 for each field or frame, and the gradation correction circuit 11 and the color correction circuit 12 of the video processing unit 10 respectively perform computations. A filter processing is performed by setting a coefficient, and a data processing section 30 is provided that outputs a filter with a response speed changed according to the feature extracted by the feature extraction section 20. The data processing unit 30 is composed of a CPU, inputs various data for video processing, generates adaptively controlled data, and applies it to the gradation correction circuit 11 and the color correction circuit 12 (hereinafter referred to as a leaky integration circuit). (Abbreviated as IIR filter) 31 and scene change analysis such as a scene change based on the features detected by the feature extraction unit 20 to obtain a fluctuation value as an index of the scene change and correspond to this fluctuation value And a coefficient changing unit 32 that determines a parameter for deriving the feedback coefficient value to change the response speed of the IIR filter 31.

  FIG. 2 is a block diagram showing a detailed configuration of the gradation correction circuit 11, and an image feature detection unit that receives an input luminance signal Y, generates histogram data based on the luminance level, and obtains average luminance and spread coefficient. 111, a plurality of gains are generated based on the histogram data, average luminance, and spread coefficient obtained by the image feature detection unit 111, and a calculation unit 112 that integrates the histogram and a value obtained by the integration are obtained. And a gradation correction unit 113 that inputs and corrects the gradation of the input luminance signal and outputs it as an output luminance signal Y ′.

  Next, the operation of the gradation correction circuit 11 will be described. As shown in FIG. 3, the image feature detection unit 111 provides a predetermined determination area fa in the screen f that is one field or one frame, and generates histogram data based on the luminance level in the determination area fa. . Further, the image feature detection unit 111 obtains an average luminance (APL) in the determination area fa. Here, the input luminance signal has 256 gradations (8 bits), and the upper 4 bits are used to generate histogram data H [i] of 16 levels (16 steps, distribution number 16).

The histogram data H [i] has different distribution spreads as shown in FIG. 4 depending on the type of input image. FIG. 4A is an example of a histogram of an image of a personal computer or the like, and the distribution is concentrated on a specific luminance level. FIG. 4B is an example of a histogram of a normal image (natural image), and 16 pieces of histogram data are completely uniformly distributed.
The difference in the distribution of the histogram data H [i] is defined as a spread coefficient expCoef, and this spread coefficient (expCoef) is calculated by the following equation.

  However, Hmax is the maximum value of H [i] (i = 0-15).

In the case of an image having a distribution as shown in FIG. 4A, the spread coefficient expCoef is 0 from (Expression 1), and in the case of an image having a completely uniform distribution as shown in FIG. 4B, the spread coefficient expCoef is (1). It becomes 600 from the formula. The spread coefficient expCoef takes a value between 0 and 600.
The histogram data H [i], average luminance, and spread coefficient obtained by the image feature detection unit 111 in FIG. The calculation unit 112 generates a plurality of gains to be described later based on the histogram data H [i], the average luminance, and the spread coefficient input in each screen f.

By the way, as shown in FIG. 4A, it is preferable not to perform video correction (image quality correction) on a personal computer image in which the luminance is unevenly distributed to a specific level. As shown in FIG. It is more desirable to perform video correction on a natural image with average brightness distribution. Therefore, the calculation unit 112 first generates a spread gain Gexp based on the spread coefficient expCoef.
FIG. 5 is an example of the spread gain Gexp based on the spread coefficient expCoef. The horizontal axis represents the spread coefficient expCoef, and the vertical axis represents the gain Gexp. As shown in FIG. 5, when the spread coefficient expCoef is an extremely small value, for example, 12 to 300, the gain Gexp is set to 1 at the maximum. The spread coefficient expCoef has a large value from 300 to 600, and the gain is reduced.

In the example shown in FIG. 5, 4, 12, and 300 are the changing points of the gain Gexp, but the present invention is not limited to this, and the gain changing point can be set arbitrarily. In this embodiment, the gain is set to 0 when the spread coefficient expCoef is from 0 to 4, but it may not be set to 0.
Next, the calculation unit 112 generates an average luminance gain Gapl based on the average luminance. FIG. 6 shows an example of the average luminance gain Gapl. The horizontal axis represents average luminance APL, and the vertical axis represents average luminance gain Gapl. As shown in FIG. 6, the average luminance gain Gapl is set to 1 at the maximum when the average luminance is low, for example, from 0 to 64, and the average luminance gain Gapl is decreased from medium luminance to high luminance, for example, from 64 to 255. The reduction start point of the average luminance gain Gapl and the ratio of the average luminance gain Gapl reduction can be arbitrarily set.

In general, when the average luminance is high, a phenomenon occurs in which the slope of the gamma curve becomes large at high luminance, becomes small at relatively low luminance, and the low luminance signal becomes darker. Since this phenomenon is not preferable, the average luminance gain Gapl has a characteristic as shown in FIG.
Further, the calculation unit 112 generates a weighting gain Gw for performing weighting on the generated histogram data H [0] to [15]. FIG. 7 shows an example of the weighting gain Gw. The horizontal axis represents the index i of the histogram data, and the vertical axis represents the weighting gain Gw. As shown in FIG. 7, for example, low luminance histogram data H [i] from i = 0 to 4 has a maximum weighting gain Gw of 1, for example, from medium luminance to high luminance where i is 5 or more. The histogram data H [i] decreases the weighting gain Gw. The reason for this characteristic is the same as the average luminance gain Gapl based on the above average luminance. The average luminance gain Gapl and the weighting gain Gw may have characteristics that gradually decrease above a predetermined luminance. Note that the reduction start point and the reduction rate can be arbitrarily set for the weighting gain Gw.

  Furthermore, the calculation unit 112 performs histogram integration as shown in FIG. 8 using the histogram data H [i] and the above-described spread gain Gexp, average luminance gain Gapl, and weighting gain Gw. The flowchart of the integration process shown in FIG. 8 is for generating points P [i] (i = 1 to 15) constituting a gradation correction curve (gamma curve) for correcting the gradation of the input luminance signal. .

In FIG. 8, first, in step S1, the integral value sum of i and point P [i] in the calculation unit 112 is set to zero. In the next step S2, it is determined whether i is less than 16. If i is 16 or more, the process proceeds to step S9. In step S3, histogram data H [i] is input. Here, since i = 0, H [0]. Subsequently, in step S4, the histogram data H [i] input in step S3 is offset using the average value Hav of H [i] (i = 1 to 15) from the calculation formula shown by the following formula (2). . The calculated value integ in step S4 is an offset value. Here, the average value Hav may be adaptively corrected with average luminance or spread coefficient.
integ = (H [i] −Hav) × 16 / Hav (2)

Next, as shown in the following equation (3) in step S5, the offset value integ obtained in step S4 is added to a predetermined fixed gain G, the spread gain Gexp already obtained from FIGS. Multiply the luminance gain Gapl and the weighting gain Gw. The fixed gain G is preset in the calculation unit 112.
integ = integ × G × Gexp × Gapl × Gw (3)

In step S6, upper limit and lower limiters are applied to the data obtained in step S5. In step S7, the data obtained in step S6 is added to the integral value sum of the point P [i]. Here, since i = 0 and sum = 0, the newly obtained sum is the data obtained in step S6. In step S8, the point P [i] of the histogram data H [i] is obtained by the following equation (4) using the integral value sum obtained in step S7.
P [i] = sum / 16 (4)

P [i] obtained here is temporarily stored in a memory (not shown) in the calculation unit 112. When step S8 ends, i is incremented by 1, and the process returns to step S2. Then, the processes from step S2 to step S8 described above are repeated until i = 15.
Through the above processing, each point from P [0] to P [15] constituting the gradation correction curve shown in FIG. 9 is generated. The horizontal axis in FIG. 9 represents the input luminance signal Y, and the vertical axis represents the output luminance signal Y ′.

If P [15] is obtained in step S8 shown in FIG. If it is determined in step S2 that i is not less than 16, the process proceeds to step S9. In step S9, a difference D between the current white level (P [15]) is obtained from a predetermined white level using the following equation (5). This is because it is necessary to correct the current white level to a predetermined white level. Here, the predetermined white level is Lw, and the predetermined black level is Lb. In this embodiment, Lb = 0.
D = Lw−Lb−P [15] (5)

In step S10, i = 0 is set again, and in step S11, it is determined whether i is less than 16. If i is smaller than 16, the process proceeds to step S12. If i is 16 or more, the process proceeds to step S13. In step S12, correction for changing the white level of each point P [i] in accordance with the difference D caused by matching the current white level P [15] with the predetermined white level Lw in step S9 is described below. (6) is used. Each point P [i] whose white level has been corrected is temporarily stored in a memory (not shown) in the calculation unit 112.
P [i] = D × (i + 1) / 16 + Lb (6)

  When the process of step S12 ends, i is incremented by 1, and the process returns to step S11. Then, the process of step S12 is repeated until i = 15. When P [15] is corrected in step S12, i becomes 16. In step S11, it is determined that i is not less than 16, and the process proceeds to step S13. In step S13, the calculation unit 112 adds new P [0] to P [15] in which the white level is corrected to the gradation correction unit 113. Note that P [0] to P [15] output in step S13 via step S12 are different from P [0] to P [15] shown in FIG. 9 obtained in step S8, but the same for convenience. The code | symbol is attached | subjected.

Next, the gradation correction unit 113 linearly interpolates P [i] (i = 1 to 15) added thereto as shown in FIG. 9, and the level of the input luminance signal Y according to the generated gradation correction curve. The tone is corrected and output as an output luminance signal Y ′. The output luminance signal Y ′ is supplied to a display device (not shown) and displayed as an image.
In this way, a gradation correction curve (gamma curve) is automatically generated for each screen f, and the input video signal is corrected according to the gradation correction curve, so that an optimum image quality can always be obtained. As a result, higher image quality can be achieved than in the conventional correction using a fixed gradation correction curve.

  By the way, in the present embodiment, the above-described video correction processing is performed on the luminance signal component of the input video signal. However, if the video correction is performed only on the luminance signal component, the color density changes, which is not preferable. It is also necessary to perform correction processing on the color signal component of the signal. Therefore, a color correction circuit 12 is provided.

  Next, the color correction circuit 12 will be described. FIG. 10 is a block diagram showing a detailed configuration of the color correction circuit 12, and an LPF (low-pass filter) 121 for removing noise from the color difference signals RY and BY and the color difference signals RY and BY. Among the coefficients calculated for each color, an angle T calculation unit 122 for calculating the angle T, correction coefficient calculation units 1231, 1232 to 123n for calculating correction coefficients for each color to be corrected using the angle T as a parameter, Adders 124 to 127 that add the same type of coefficients, hue / saturation correction processing unit 128 that corrects the color difference signals RY and BY based on the outputs of the adders 124 and 125, and adders 126 and 127 A luminance gamma processing / saturation gamma processing unit 129 that corrects luminance gamma and saturation gamma by inputting the output and input luminance signal Y, and corrected input luminance signal Y, color difference signals RY, BY Saturation And a saturation correction unit 130 that corrects based hue correction coefficient SatCorrect.

Next, the operation of the color correction circuit 12 shown in FIG. 10 will be described. The color difference signals RY and BY are added to the angle T calculation unit 122 via the LPF 121 for noise removal, and the angle T calculation unit 122 calculates the angle T of the color difference signals RY and BY. Here, for example, as shown in FIG. 11, the color difference signals RY and BY are represented by a color difference plane in which BY is an X axis (horizontal axis) and RY is a Y axis (vertical axis). In this color difference plane, the rotation direction represents hue, and the radial direction represents saturation. Further, as shown in FIG. 11A, L0, L1, and L2 at angles θ ₀ , θ ₁ , and θ ₂ (θ ₁ <θ ₀ <θ ₂ ) represent equal hue lines. The equi-hue line L0 is a correction center line and is set by the correction center angle θ ₀ , and the angles θ _{1 to} θ ₂ are correction regions.
Correction coefficient calculation units 1231, 1232 to 123n (hereinafter, arbitrary correction coefficient calculation units are 123i: i = 1, 2 to n) are provided according to the number of colors to be corrected, and are corrected using the angle T as a parameter. A saturation correction coefficient Si, a Y (luminance) gamma correction coefficient GYi, and a C (saturation) gamma correction coefficient Gγi are calculated according to the number of target colors. At this time, in order to change the correction coefficient R for determining the predetermined correction center angles θ ₀ ′, θ _{1 to} θ ₂ corresponding to the color to be corrected, and the degree of correction, for each display, with respect to the correction coefficient calculation unit 123 i Hue 1 coefficient p1 ′ for correcting the hue correction coefficient Pi, Hue 2 coefficient p2 ′, a saturation coefficient s ′ for correcting the saturation correction coefficient Si, and a Y gamma coefficient gy for correcting the Y gamma correction coefficient GYi. ', C gamma coefficient gc' for correcting C gamma correction coefficient GCi is set, and correction coefficient calculation unit 123i calculates correction coefficients Si, Pi, GYi, GCi using these coefficients. Since this calculation process is shown in Patent Document 2, detailed description thereof is omitted.

(1) In “hue 1 correction”, as shown in FIG. 11A, the hue in L1 to L2 is corrected to the center hue L0 (or L1, L2) centered on the equihue line L0. Thereby, for example, the color near the skin color is brought close to the skin color.
(2) In “hue 2 correction”, as shown in FIG. 11B, the hues in L1 to L2 are rotated and corrected in the same direction.
(3) In “saturation correction”, the radial direction of the colors in L1 to L2 is corrected as shown in FIG.
(4) In “Y gamma correction”, the gamma of the Y signal in the colors within L1 to L2 is corrected as in FIG.
(5) In “C gamma correction”, saturation gamma is corrected in conjunction with “Y gamma correction”.
(6) “Saturation correction at high saturation” suppresses hue fluctuation due to saturation at the time of increase in saturation for a high saturation color. Corrections (1) to (5) are performed by correction coefficient calculation units 1231 and 1232 to 123n shown in FIG.

  Returning to FIG. 10, the adder 124 adds the saturation correction coefficients Si calculated by the correction coefficient calculation unit 123 i to obtain the sum ΣSi, and the adders 125 to 127 respectively add the sum ΣPi of the hue correction coefficients Pi, respectively. A total ΣGYi of Y gamma correction coefficients GYi and a total ΣGCi of C gamma correction coefficients GCi are obtained. The hue / saturation correction processing unit 128 corrects the saturation and hue of the color difference signals RY and BY by performing a product-sum operation based on the saturation correction coefficient total value ΣSi. FIG. 10 shows correction calculation colors in which the color difference signals RY and BY are iri and iby, respectively. The luminance gamma processing / saturation gamma processing unit 129 converts the luminance gamma of the Y signal and the saturation gammas of the color difference signals RY and BY corrected by the hue / saturation correction processing unit 128 into Y gamma correction coefficients. Correction is performed based on the total value ΣGYi and the C-gamma correction coefficient total value ΣGCi. Further, since the saturation is apparently changed when the saturation gamma is corrected, the luminance correction unit 130 corrects the luminance signal Y and the color difference signal R corrected by the luminance gamma processing / saturation gamma processing unit 129 in order to prevent this. The saturation of −Y and BY is corrected based on the correction coefficient SatCorrect.

  In the color correction circuit 12 shown in FIG. 10, the correction coefficient calculation unit 123i has a hue pi 'hue 2' coefficient pi 'hue 2' for correcting the hue correction coefficient Pi in order to change the degree of correction for each display. The coefficient p2 ′, the s ′ for correcting the saturation correction coefficient Si, the Y gamma coefficient gy ′ for correcting the Y gamma correction coefficient GYi, and the C gamma coefficient gc ′ for correcting the C gamma correction coefficient GCi are irrespective of the image. Since it is uniquely set, in the case of an overall dark scene with a poor S / N ratio, the noise is also corrected and the noise becomes conspicuous. In order to solve such problems, a feature extraction unit 20 and a data processing unit 30 are provided.

  The feature extraction unit 20 is configured by hardware capable of high-speed processing in order to process information in pixel units of the video signal. The feature extraction unit 20 detects a histogram value histBlock, an average luminance value APL, a minimum luminance value MNI, a maximum luminance value MAX, and a motion change amount Move as image features from the input signal and supplies them to the data processing unit 30.

The data processing unit 30 requires analysis for each frame and complicated arithmetic processing, and the functions of the IIR filter 31 and the coefficient changing unit 32 are realized by software using a CPU. Among these, the IIR filter 31 is an integrator with a leak, and is configured as shown in FIG. That is, an adder 311 that adds the input X and the feedback value, a multiplier 312 that multiplies the output of the adder 311 by a constant k, and an output of the multiplier 312 is delayed and added to the adder 311 as a feedback signal. The delay element 313 includes a multiplier 314 that multiplies the output of the adder 311 by a coefficient of 1/2 ⁿ and outputs Y as a positive number that can be changed from the outside of the filter. This IIR filter 31 adds, as an input X, a hue 1 coefficient p1, a hue 2 coefficient p2, a saturation coefficient s, a Y gamma coefficient gy, and a C gamma coefficient gc set inside or outside the data processing unit 30. As an output Y, a hue 1 coefficient p1 ′, a hue 2 coefficient p2 ′, a saturation coefficient s ′, a Y gamma coefficient gy ′, and a C gamma coefficient gc ′ to be applied to the color correction circuit 12 are obtained.

The IIR filter 31 varies the response speed by changing the value of the parameter n for deriving the feedback coefficient value, and has the following characteristics.
(A) Large suppression effect with a small number of parts (b) Easy response characteristic change (c) A response without hunting is obtained, and the transfer characteristic is expressed as a first-order lag characteristic by the following equation (7)

However, k is a feedback coefficient value (leak coefficient) (k <1).

FIG. 13 is a characteristic diagram of the IIR filter 31 described above. The horizontal axis represents the number of fields, and the vertical axis represents the transfer characteristics. Here, the response speed can be easily varied by the value of the parameter n for deriving the feedback coefficient value (leak coefficient).
The coefficient changing unit 32 determines the parameter n, inputs the image features supplied from the feature extracting unit 20, analyzes scene changes such as scene changes, and is an index of scene changes. Find the variation. An example of calculating the fluctuation value is shown below.

However,
histBlock new: Current histogram value
histBlock: Histogram value before 1 process (frame)
APL new: Current average luminance value
APL: Average luminance value before 1 process (frame)
MIN new: Current minimum luminance value
MIN: Minimum luminance value before processing (frame)
MAX new: Current maximum brightness value
MAX: Maximum brightness value before processing (frame)
Move: The amount of change in motion before the current process and one process (frame). Thereby, based on various image information, the amount of change between the present and one process (frame) before is obtained. In this example, all the extracted image information is used. However, the present invention is not limited to this, and the variation value may be calculated by combining several pieces of image information, even if it is one of the extracted image information. .

  The coefficient changing unit 32 changes the parameter n for deriving the feedback coefficient value of the IIR filter 31 by the fluctuation value obtained by the equation (8) for each field or frame. In this case, when the variation value is large, it is determined that the scene change is a scene change, and the response speed is increased by decreasing the value of the parameter n. On the other hand, when the variation value is small, it is determined that there is no significant change in the scene, and the parameter n is increased to provide a suppression effect against minute variations. Hereinafter, a specific example in which the value of the parameter n is changed will be described.

  FIGS. 14A and 14B are examples of changes in the histogram, such as a scene change in which the scene suddenly changes from a dark image to a bright image. The horizontal axis represents the gradation and the vertical axis represents the frequency. Yes. In such a case, since the fluctuation value becomes a large value, it is determined that the scene has changed greatly, and the value of the parameter n of the IIR filter 31 is decreased to increase the response speed.

  FIG. 15 is a diagram showing the relationship between the fluctuation value obtained from various image information and the parameter n for deriving the feedback coefficient value of the IIR filter 31 in order to determine the value of the parameter n. The vertical axis represents the parameter value. Here, s0 is set to the lower limit of the range in which the fluctuation value normally changes, s1 is set to the upper limit, and the value of parameter n is set to “8” in the range where the fluctuation value is smaller than the set value s0. In the range where the fluctuation value is larger than the set value s1, the value of the parameter n is set to "4", and in the range where the fluctuation value is larger than s0 and smaller than s1, the value of the parameter n becomes smaller as the fluctuation value becomes larger. Is linearly changed from “8” to “4”. In this case, the set values s0 and s1 can be appropriately changed, and the state of change in the range where the set value is larger than s0 and smaller than s1 is not limited to a straight line, and the value of n changes monotonously from a large value to a small value. You can also do it.

  In the data processing unit 30, various data for image processing, that is, as described above, the hue 1 coefficient p1, the hue 2 coefficient p2, the saturation coefficient s, the Y gamma coefficient gy, and the C gamma coefficient gc are stored. These data are passed through the IIR filter 31, and the tone correction circuit 11 converts the hue 1 ′ coefficient p1 ′, the hue 2 ′ coefficient p2 ′, the saturation coefficient s ′, the Y gamma coefficient gy ′, and the C gamma coefficient gc ′. Added to.

  In this way, the characteristics of the input video signal are detected for each field or each frame, and scene changes such as scene changes are analyzed based on these characteristics to obtain the fluctuation values of the scene changes, and the fluctuation values are large. In this case, the parameter n for deriving the feedback coefficient value of the IIR filter 31 is decreased to increase the response speed. Conversely, when the fluctuation value is small, the parameter n for deriving the feedback coefficient value of the IIR filter 31 is increased to increase the suppression effect. Since the coefficient value processed by the color correction circuit 12 is optimally controlled, the correction target color can be corrected vividly and noise can be prevented from being noticeable even in a dark scene. Can do.

  In the above embodiment, the hue 1 ′ coefficient p1 ′, the hue 2 ′ coefficient p2 ′, and the saturation correction coefficient Si that are added to the correction coefficient calculators 1231, 1232-123n are corrected by s ′, Y gamma. Only the Y gamma coefficient gy ′ for correcting the correction coefficient GYi and the C gamma coefficient gc ′ for correcting the C gamma correction coefficient GCi were supplied with the response speed of the IIR filter 31 changed. For the histogram data H [i] for which the image data is being processed, the IIR filter 31 is interposed, and the image quality can be optimally corrected even in a scene change in which the histogram data changes greatly by changing the response speed.

  In addition, the coefficient supplied by changing the response speed of the IIR filter 31 may be a part of the above-mentioned coefficient, or may be supplied including a coefficient other than the above-described coefficient. Either one or both of the tone correction circuit 11 and the color correction circuit 12 may be used.

It is a block diagram which shows the structure of one Embodiment of the video signal processing apparatus which implements the video signal processing method which concerns on this invention. It is a block diagram which shows the detailed structure of the gradation correction circuit which comprises one Embodiment. It is a figure which shows the determination area | region in 1 screen, in order to demonstrate the operation | movement of one Embodiment. It is the figure which showed the example of distribution of histogram data in order to demonstrate operation | movement of one Embodiment. FIG. 6 is a diagram illustrating an example of a spread gain based on a spread coefficient of histogram data in order to explain the operation of an embodiment. It is the figure which showed an example of the average luminance gain based on average luminance, in order to demonstrate operation | movement of one Embodiment. In order to demonstrate operation | movement of one Embodiment, it is the figure which showed an example of the weighting gain based on the weight of histogram data. FIG. 6 is a flowchart for generating values constituting a gradation correction curve in order to explain the operation of an embodiment. FIG. 5 is a diagram illustrating a method for generating a gradation correction curve in order to explain the operation of an embodiment. It is a block diagram which shows the detailed structure of the color correction circuit which comprises one Embodiment. FIG. 4 is an explanatory diagram showing processing of a color correction circuit in order to explain the operation of an embodiment. It is a block diagram which shows the structure of the IIR filter which comprises one Embodiment. It is a characteristic view of the IIR filter which comprises one Embodiment. It is a figure showing the example of change of a histogram, such as a sea change where a scene changes suddenly, in order to explain operation of one embodiment. FIG. 4 is a diagram showing the relationship between a fluctuation value and a parameter n for deriving a feedback coefficient value of an IIR filter in order to explain the operation of an embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image processing part 11 Gradation correction circuit 12 Color correction circuit 20 Feature extraction part 30 Data processing part 31 Leak type integration circuit (IIR filter)
32 coefficient changing unit 111 image feature detecting unit 112 calculating unit 113 gradation correcting unit 121 LPF (low-pass filter)
122 Angle T Calculation Unit 124 to 127 Adder 128 Hue / Saturation Correction Processing Unit 129 Luminance Gamma Processing / Saturation Gamma Processing Unit 130 Saturation Correction Unit 1231 to 123n Correction Coefficient Calculation Unit

Claims (2)

In a video signal processing method for correcting gradation and color signal components of a video signal,
Supplying a parameter used for correcting at least one of the gradation and the color signal component by changing a response speed by changing a feedback coefficient value of a leaky integration circuit;
Extracting at least one of a histogram, an average luminance value, a minimum luminance value, a maximum luminance value, and a motion change amount as a feature of the video signal for each field or each frame;
Obtaining a variation value that is an index of a scene change based on the extracted feature, and setting a feedback coefficient value of the leaky integration circuit according to the variation value for each field or each frame,
The step of setting the feedback coefficient value changes and sets the feedback coefficient value in a direction in which the response speed increases as the fluctuation value increases within a predetermined range in which the fluctuation value changes. Processing method. In a video signal processing apparatus comprising a gradation correction circuit for correcting the gradation of a video signal and a color correction circuit for correcting a color signal component,
A leak type integration circuit that inputs parameters used for correcting at least one of the gradation and the color signal component, and supplies the parameters while changing the response speed;
Feature extraction means for extracting at least one of a histogram, an average luminance value, a minimum luminance value, a maximum luminance value, and a motion change amount as a feature of the video signal for each field or each frame;
Based on the features extracted by the feature extraction means, a variation value that is an index of scene change is obtained, and a coefficient change that sets the feedback coefficient value of the leaky integration circuit for each field or frame according to the variation value Means and
The video signal processing apparatus, wherein the coefficient changing means changes and sets the feedback coefficient value in a direction in which the response speed increases as the fluctuation value increases within a predetermined range in which the fluctuation value changes.

Images