WO2022244073A1 - Image processing device, program, and image processing method - Google Patents

Abstract

This image processing device (100) comprises: a feature quantity extraction unit (101) for extracting a feature quantity from an input image signal; a grayscale correction table creation unit (102) for creating, using the feature quantity, a grayscale correction table that indicates a correspondence relationship between an input grayscale that corresponds to the feature quantity and an output grayscale; a maximum value detection unit (103) for detecting, for each pixel, the maximum value of grayscale from the input image signal; a multiplication value calculation unit (105) for calculating, from the grayscale correction table, a multiplication value for the maximum value in order to make the maximum value an output grayscale that corresponds to the maximum value; a minimum value detection unit (104) for detecting, for each pixel, the minimum value of grayscale from the input image signal; a multiplication value correction unit (206) for employing the multiplication value as a corrected multiplication value when the multiplication value is equal to or greater than 1, and employing a value obtained by correcting the multiplication value so as to become closer to 1 commensurately with a greater difference between the maximum and minimum values as a corrected multiplication value when the multiplication value is less than 1; and multipliers (107A, 107B, 107C) for multiplying the corrected multiplication value by the input image signal.

Description

Image processing device, program and image processing method

The present disclosure relates to an image processing device, a program, and an image processing method.

With the growing interest in environmental issues, the need for power saving is increasing. In an image display device with a backlight as a light source, such as a liquid crystal display device, the brightness of the display itself is maintained by dimming the brightness of the backlight and increasing the light transmittance of the liquid crystal panel through image processing. power can be saved.

However, in self-luminous devices such as LEDs including micro LEDs (Light Emitting Diodes) and mini LEDs, or organic ELs (Electro Luminescence), if the brightness is dimmed for power saving, the display will be dark and the image quality will deteriorate. expected to decline.

Therefore, self-luminous devices maintain the brightness of bright areas and darken the brightness of dark areas to increase contrast and maintain image quality. However, when an image with high saturation becomes dark, the vividness of colors may be lost and the image quality may deteriorate.

The image display device described in Patent Document 1 converts the lower the luminance value of each pixel, excluding the pixel with the lowest luminance value, into the pixel with the lowest luminance value with a higher frequency. and achieve power saving. By doing so, relatively dark pixels are displayed darker, and visibility for the user can be maintained.

Japanese Patent No. 5226188

The purpose of conventional image display devices is to save power by not changing to a color different from the original display color of the image, such as by changing to a color that consumes less power.

However, even if the display color of the original image is not changed, darkening the brightness of the highly saturated image will lead to loss of vividness and deterioration of image quality. For example, if the brightness of a red pixel is reduced too much, it will appear brown. This is because even if there is no change in chromaticity, or there is almost no change in chromaticity, the appearance of color changes due to a change in lightness.

One or more aspects of the present disclosure aim to achieve power saving without impairing color vividness and user visibility.

An image processing apparatus according to an aspect of the present disclosure includes a feature amount extraction unit that extracts a predetermined feature amount for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. Then, the number of pixels is totaled for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division with the smallest feature amount for normalization, thereby obtaining a gradation corresponding to the feature amount. a gradation correction information creating unit that creates gradation correction information indicating a correspondence relationship between an input gradation and an output gradation that is a gradation obtained by correcting the input gradation; a maximum value detection unit for detecting the maximum value of the gradation; and calculating a multiplication value of the maximum value from the gradation correction information so as to make the maximum value the output gradation corresponding to the maximum value. a multiplication value calculation unit, a minimum value detection unit for detecting the minimum value of the gradation for each pixel from the input image signal, and if the multiplication value is 1 or more, the multiplication value is used as a correction multiplication value. , when the multiplication value is less than 1, the multiplication value is corrected so that the multiplication value approaches 1 as the difference between the maximum value and the minimum value increases, and the modified multiplication value is set as the multiplication value. and a multiplier for multiplying the input image signal by the modified multiplication value.

An image processing device according to an aspect of the present disclosure is a predetermined feature amount for each predetermined region from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. an area feature amount calculation unit that calculates an area feature amount; a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel; and a gradation corresponding to the synthesized feature amount. a gradation correction information storage unit for storing a plurality of gradation correction information indicating a correspondence relationship between an input gradation and each output gradation which is a gradation obtained by correcting the input gradation; and the plurality of gradation corrections. a gradation correction information selection unit that selects gradation correction information corresponding to the synthetic feature amount for each pixel from the information; and a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal. and multiplication value calculation for calculating a multiplication value for the maximum value in order to make the maximum value the output gradation corresponding to the maximum value for each pixel from the gradation correction information selected for each pixel. a minimum value detection unit for detecting the minimum value of the gradation for each pixel from the input image signal; when the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value; a multiplied value correcting unit that, when the value is less than 1, sets the corrected multiplied value to a value obtained by correcting the multiplied value so that the larger the difference between the maximum value and the minimum value, the closer to 1; and a multiplication unit that multiplies the input image signal by the modified multiplication value.

An image processing apparatus according to an aspect of the present disclosure includes a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color, a feature amount extracting unit for extracting a predetermined feature amount for each pixel from an input image signal composed of a color difference signal indicating the color difference of the feature amount; By normalizing by adding the number of pixels in descending order of the amount of the division, an input gradation corresponding to the feature quantity and an output gradation obtained by correcting the input gradation are obtained. a gradation correction information creation unit that creates gradation correction information indicating the correspondence relationship between; a reference value specifying unit that specifies a reference value for referring to the gradation correction information for each pixel from the input image signal; a multiplication value calculation unit for calculating a multiplication value for the reference value in order to set the reference value to the output gradation corresponding to the reference value from the gradation correction information; a maximum value specifying unit for specifying the maximum absolute value of the color difference; when the multiplication value is 1 or more, the multiplication value is a modified multiplication value; and when the multiplication value is less than 1, A multiplied value correcting unit that modifies the multiplied value so that it approaches 1 as the maximum value increases, and a multiplication unit that multiplies the input image signal by the modified multiplied value. characterized by

An image processing apparatus according to an aspect of the present disclosure includes a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color, an area feature amount calculation unit for calculating an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a color difference signal indicating the color difference of each pixel; a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the input gradation corresponding to the synthesized feature amount; and the output gradation that is the corrected gradation of the input gradation a gradation correction information storage unit for storing a plurality of gradation correction information indicating a correspondence relationship between the gradation correction information and a gradation correction information storage unit for selecting gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information; a tone correction information selection unit; a reference value identification unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal; and the tone correction information selected for each pixel, a multiplication value calculation unit for calculating a multiplication value for the reference value in order to set the reference value to the output gradation corresponding to the reference value for each pixel; a maximum value specifying unit for specifying a maximum absolute value; when the multiplication value is 1 or more, the multiplication value is a modified multiplication value; and when the multiplication value is less than 1, the maximum value and a multiplication unit that multiplies the input image signal by the modified multiplication value, and a multiplication unit that modifies the multiplication value so that it approaches 1 as the value increases. do.

A program according to an aspect of the present disclosure causes a computer to perform feature amount extraction for extracting a predetermined feature amount for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. In part, the number of pixels is totaled for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division with the smallest feature amount for normalization, thereby obtaining a gradation corresponding to the feature amount. a gradation correction information creating unit that creates gradation correction information indicating a correspondence relationship between an input gradation and an output gradation that is a gradation obtained by correcting the input gradation; a maximum value detection unit for detecting a maximum value of gradation; and multiplication for calculating a multiplication value for the maximum value in order to make the maximum value the output gradation corresponding to the maximum value from the gradation correction information. a value calculation unit, a minimum value detection unit for detecting the minimum value of the gradation for each pixel from the input image signal; a multiplication value correction unit that, when the value is less than 1, sets the corrected multiplication value to a value obtained by correcting the multiplication value so that the larger the difference between the maximum value and the minimum value, the closer to 1; and a multiplier for multiplying the input image signal by the corrected multiplication value.

A program according to an aspect of the present disclosure causes a computer to use a predetermined feature amount for each predetermined region from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. an area feature amount calculation unit that calculates a certain area feature amount; a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel; a gradation correction information storage unit for storing a plurality of gradation correction information indicating a correspondence relationship between a gradation and each output gradation which is a gradation obtained by correcting the input gradation; , a gradation correction information selection unit that selects gradation correction information corresponding to the combined feature amount for each pixel; a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal; a multiplication value calculation unit for calculating a multiplication value for the maximum value in order to set the maximum value to the output gradation corresponding to the maximum value for each pixel from the gradation correction information selected by the input; a minimum value detection unit for detecting the minimum value of the gradation for each pixel from an image signal; when the multiplication value is 1 or more, the multiplication value is a modified multiplication value; and the multiplication value is less than 1. a multiplied value correction unit that sets the corrected multiplied value to a value obtained by correcting the multiplied value so that the larger the difference between the maximum value and the minimum value, the closer to 1, and the corrected multiplied value It is characterized by functioning as a multiplication unit that multiplies the input image signal.

A program according to an aspect of the present disclosure causes a computer to generate a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color. A feature amount extraction unit for extracting a predetermined feature amount for each pixel from an input image signal composed of a color difference signal indicating a color difference between the feature amount and the By normalizing by adding the number of pixels in descending order of the amount of the division, an input gradation corresponding to the feature quantity and an output gradation obtained by correcting the input gradation are obtained. a reference value specifying unit for specifying a reference value for referring to the tone correction information for each pixel from the input image signal; a multiplication value calculation unit for calculating a multiplication value for the reference value in order to set the reference value to the output gradation corresponding to the reference value from the tone correction information; A maximum value specifying unit for specifying the maximum value of absolute values, when the multiplication value is 1 or more, the multiplication value is a modified multiplication value, and when the multiplication value is less than 1, the maximum value is It functions as a multiplied value correcting unit that corrects the multiplied value so that the multiplied value approaches 1 as the multiplied value increases, and the multiplied value that multiplies the input image signal by the corrected multiplied value. do.

A program according to an aspect of the present disclosure causes a computer to generate a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color. an area feature amount calculation unit for calculating an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a color difference signal indicating a color difference between and, for each pixel, the area feature amount By synthesizing, a feature amount synthesizing unit that calculates a synthesized feature amount, an input tone that is a tone corresponding to the synthesized feature amount, and an output tone that is a tone obtained by correcting the input tone, and a gradation correction information storage unit for storing a plurality of gradation correction information indicating a correspondence relationship between the gradation correction information, and a gradation correction that selects gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information an information selection unit; a reference value identification unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal; a multiplication value calculation unit for calculating a multiplication value for the reference value in order to set the reference value to the output gradation corresponding to the reference value; a maximum absolute value of the color difference for each pixel from the input image signal when the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the greater the maximum value, the closer to 1 and a multiplication unit that multiplies the input image signal by the modified multiplication value.

An image processing method according to an aspect of the present disclosure extracts a predetermined feature amount for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel, and extracts the feature amount is divided in ascending order, and normalization is performed by adding the number of pixels in descending order of the feature amount to obtain the input gradation corresponding to the feature amount. creating gradation correction information indicating a correspondence relationship with an output gradation which is a gradation obtained by correcting the input gradation; detecting the maximum value of the gradation for each pixel from the input image signal; From the tone correction information, a multiplication value for the maximum value is calculated in order to set the maximum value to the output gradation corresponding to the maximum value, and from the input image signal, the minimum value of the gradation is calculated for each pixel. and if the multiplication value is greater than or equal to 1, the multiplication value is taken as a modified multiplication value; if the multiplication value is less than 1, the difference between the maximum value and the minimum value is large. A modified multiplication value is obtained by modifying the multiplication value so as to approach 1 as much as possible, and the input image signal is multiplied by the modification multiplication value.

An image processing method according to an aspect of the present disclosure is a predetermined feature amount for each predetermined region from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. calculating an area feature amount, synthesizing the area feature amount for each pixel to calculate a synthesized feature amount, and correcting the input gradation that is the gradation corresponding to the synthesized feature amount and the input gradation; a plurality of gradation correction information indicating a correspondence relationship with each output gradation, which is the gradation obtained by the gradation; and detecting the maximum value of the gradation for each pixel from the input image signal, and determining the maximum value corresponding to the maximum value for each pixel from the gradation correction information selected for each pixel. In order to obtain an output gradation, a multiplication value for the maximum value is calculated, the minimum value of the gradation is detected for each pixel from the input image signal, and when the multiplication value is 1 or more, the A multiplied value is a modified multiplied value, and when the multiplied value is less than 1, the modified multiplied value is adjusted so as to approach 1 as the difference between the maximum value and the minimum value increases. A multiplication value is used, and the input image signal is multiplied by the modified multiplication value.

An image processing method according to an aspect of the present disclosure includes a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color. extracting a predetermined feature amount for each pixel from an input image signal composed of a color difference signal indicating the color difference of , counting the number of pixels for each section divided in ascending order of the feature amount; By normalizing by adding the number of pixels in the order of , the correspondence relationship between the input gradation that is the gradation corresponding to the feature amount and the output gradation that is the corrected gradation of the input gradation is shown. gradation correction information is created, a reference value for referring to the gradation correction information is specified for each pixel from the input image signal, and the reference value is associated with the reference value from the gradation correction information In order to obtain the output gradation, a multiplied value for the reference value is calculated, the maximum value of the absolute value of the color difference is specified for each pixel from the input image signal, and if the multiplied value is 1 or more, sets the multiplied value as a modified multiplied value, and if the multiplied value is less than 1, sets the modified multiplied value to a value obtained by modifying the multiplied value so as to approach 1 as the maximum value increases; The input image signal is multiplied by the multiplication value.

An image processing method according to an aspect of the present disclosure includes a luminance signal indicating luminance of a first color, the first color, and second and third colors different from the first color. From the input image signal consisting of the color difference signal indicating the color difference of, for each predetermined area, a region feature amount, which is a predetermined feature amount, is calculated, and for each pixel, by synthesizing the area feature amount, A plurality of gradation corrections indicating a correspondence relationship between an input gradation that is a gradation corresponding to the synthetic feature amount and each output gradation that is a gradation obtained by correcting the input gradation. To store information, select tone correction information corresponding to the composite feature amount for each pixel from the plurality of tone correction information, and refer to the tone correction information for each pixel from the input image signal. is specified, and from the gradation correction information selected for each pixel, in order to make the reference value the output gradation corresponding to the reference value, the multiplication value for the reference value is and specifying the maximum value of the absolute value of the color difference for each pixel from the input image signal, and if the multiplication value is 1 or more, the multiplication value is set as a modified multiplication value, and the multiplication value is 1. If the maximum value is less than 1, the multiplied value is corrected to be closer to 1 as the maximum value increases, and the corrected multiplied value is multiplied by the input image signal.

According to one or more aspects of the present disclosure, power saving can be achieved without impairing color vividness and user visibility.

1 is a block diagram schematically showing the configuration of an image processing apparatus according to Embodiment 1; FIG. It is an example of a histogram for creating a gradation correction table. 4 is a graph showing curves corresponding to tone correction tables; It is a graph which shows the example which corrected both the maximum value of a slope, and the minimum value. 4 is a graph for explaining a method of calculating a multiplication value in Embodiment 1; (A) and (B) are block diagrams showing hardware configuration examples. 4 is a flowchart showing operations in the image processing apparatus according to Embodiment 1; 2 is a block diagram schematically showing the configuration of an image processing apparatus according to Embodiment 2; FIG. FIG. 12 is a block diagram schematically showing the configuration of an image processing apparatus according to Embodiment 3; FIG. (A) and (B) are schematic diagrams for explaining a process of calculating a synthetic feature amount. Fig. 4 is a graph showing weights corresponding to horizontal and vertical positions; FIG. 4 is a schematic diagram showing a graph plotting a plurality of gradation correction tables; FIG. 12 is a block diagram schematically showing the configuration of an image processing apparatus according to Embodiment 4; FIG. 1 is a block diagram schematically showing a first configuration example of an image display device; FIG. FIG. 4 is a block diagram schematically showing a second configuration example of the image display device;

Embodiment 1.
FIG. 1 is a block diagram schematically showing the configuration of an image processing apparatus 100 according to Embodiment 1. As shown in FIG.
The image processing device 100 is a device that processes an image. The image may be still or moving, and may or may not be accompanied by sound. In particular, a moving image is also referred to as an image, but is referred to as an image here.
The image processing apparatus 100 includes a feature amount extraction unit 101, a tone correction table generation unit 102, a maximum value detection unit 103, a minimum value detection unit 104, a multiplication value calculation unit 105, a multiplication value correction unit 106, It has multipliers 107A, 107B and 107C.

The feature quantity extraction unit 101 extracts a predetermined feature quantity for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel.
In Embodiment 1, the feature amount extraction unit 101 extracts feature amounts from the color signals RIN, GIN, and BIN forming the input image signal. For example, the feature amount extraction unit 101 calculates feature amounts such as brightness or lightness from the color signals RIN, GIN, and BIN.
Specifically, images are generally composed of the three primary colors of red, green, and blue. The input image signal is composed of a color signal RIN indicating red gradation, a color signal GIN indicating green gradation, and a color signal BIN indicating blue gradation. Luminance is obtained by adding red, green, and blue in a certain ratio, and there is a general formula, so the explanation is omitted here. Brightness can be the maximum value of red, green or blue at each pixel.

Note that FIG. 1 omits a delay unit for delaying the input image signal. The image processing apparatus 100 is assumed to include such a delay section to delay the input image signal.

The gradation correction table creation unit 102 creates a table for correcting the gradation of the color signals RIN, GIN, and BIN, which are input image signals, using the feature amount extracted by the feature amount extraction unit 101 .
For example, the gradation correction table creation unit 102 counts the number of pixels for each section divided in descending order of the feature amount, adds the number of pixels in the order of the smallest feature amount, and normalizes to correspond to the feature amount. a gradation correction information creating unit that creates a gradation correction table, which is gradation correction information indicating the correspondence relationship between an input gradation that is a gradation to be corrected and an output gradation that is a gradation obtained by correcting the input gradation; .
Processing in the gradation correction table generation unit 102 will be described below.

FIG. 2 is an example of a histogram for creating a gradation correction table.
The gradation correction table creation unit 102 divides the feature amount such as brightness or lightness of all pixels included in one image into 16, and counts which divided area each of all pixels falls into. , to get the histogram. Here, the 16-divided regions are the divisions. For example, in the example of FIG. 2, the first segmented region of 0≦feature amount≦64, the second segmented region of 65≦feature amount≦128, the third segmented area of 129≦feature amount≦192, and the third segmented area of 129≦feature amount≦193≦feature fourth segmented region with amount≦256, fifth segmented region with 257≦feature amount≦320, sixth segmented area with 321≦feature amount≦384, seventh segmented area with 385≦feature amount≦448, 449 8th segmented region of ≤feature value≤512, 9th segmented region of 513≤feature value≤576, 10th segmented region of 577≤feature value≤640, 11th segment of 641≤feature value≤704 12th divided region with 705≦feature amount≦768 13th divided area with 769≦feature amount≦832 14th divided area with 833≦feature amount≦896 897≦feature amount≦960 and the 16th segmented region of 961≦feature amount≦1023. The frequency here is the number of pixels.

Although the feature amount is divided into 16 here, the number of divisions is not limited to 16. For example, the division number may be 32, 8, or 10. As the number of divisions increases, the accuracy of gradation correction increases and the circuit size increases. If the number of divisions is a power of 2, the digital operation may be simplified by bit shifting.

Next, the gradation correction table creation unit 102 calculates the slope proportional to the frequency included in each divided area, and specifies the linear equation for each divided area.
For example, in the case of the i-th divided area, the equation of the straight line can be specified by the following equation (1).
y=a _i x+b _i−1 (1)
Note that i is an integer that satisfies 0≦i<N, and N is the number of divisions.

Here, y indicates the output gradation, and x is the feature amount corresponding to the input gradation.
Also, a _i is the slope, which is normalized by the number of pixels j of the image to be processed, as in the following equation (2), for example.
a _i =d _i ×N÷j (2)
where d _i is the frequency of the i-th segmented region.

In addition, b _i−1 is the maximum value in the i−1 th division area calculated by the equation of the straight line in the i−1 th division area. Note that when i=0, b _i−1 becomes “0”.

FIG. 3 is a graph showing a curve (here, a polygonal line) connecting the values calculated by the above straight line equation.
Then, the gradation correction table creation unit 102 corrects the slope _ai in the generated straight line equation.
For example, the gradation correction table creation unit 102 corrects at least one of the maximum value and minimum value of the slope _ai . Specifically, the tone correction table creation unit 102 may limit the maximum value of the slope _ai . In this case, the gradation correction table creation unit 102 makes the maximum value of the slope _ai smaller. Any method such as subtracting a predetermined value or multiplying by a predetermined decimal number of 1 or less can be adopted as a method for reducing the value. Also, the gradation correction table creation unit 102 increases the minimum value of the slope _ai . As for the method of increasing, any method such as adding a predetermined value or multiplying by a predetermined number of 1 or more can be adopted.

FIG. 4 is a graph showing an example in which both the maximum and minimum values of the slope _ai are corrected.
In FIG. 4, the solid line indicates the curve after correction, and the dashed line indicates the curve before correction.
As shown in FIG. 4, the slope in the divided area of gradation 640 to 704 is the maximum value, this slope is corrected to a smaller value, and the slope in the divided area of gradation 704 to 768 is the minimum value. value, and this slope has been corrected to a larger value.

Also, the correction by the gradation correction table creation unit 102 is not limited to the example described above. For example, the gradation correction table creating unit 102 can set a limit for each divided area. For example, by setting the minimum value of the slope of the high gradation portion above a predetermined gradation to 1 and the maximum value of the slope of the low gradation portion below the predetermined gradation to 1, the curve can be When looking at the gradation area, it becomes convex downward, the output gradation becomes smaller than the input gradation, and the power consumption is reduced. In addition, it is generally known that since the gradient of the input/output curve in a gradation area with a large number of distributions in the histogram becomes large, the difference in gradation becomes large in many pixels in the image, which has the effect of increasing the contrast. ing.

Then, the gradation correction table creation unit 102 creates a gradation correction table showing the correspondence between the input gradation and the output gradation at the break point of the curve corrected as described above.

Returning to FIG. 1, the maximum value detection unit 103 detects the maximum value of gradation for each pixel from the input image signal.
For example, the maximum value detection unit 103 detects the maximum value of gradation in each pixel of the input image indicated by the color signals RIN, GIN, and BIN. The detected maximum value is provided to multiplication value calculation section 105 and multiplication value correction section 106 . For example, when the pixel value of a certain pixel is (R, G, B)=(0, 409, 818), "818" is the maximum value.

The minimum value detection unit 104 detects the minimum value of gradation for each pixel from the input image signal.
For example, the minimum value detection unit 104 detects the minimum gradation value of each pixel of the input image indicated by the color signals RIN, GIN, and BIN. The detected minimum value is provided to multiplication value correction section 106 . For example, when the pixel value of a certain pixel is (R, G, B)=(0, 409, 818), "0" is the minimum value.

A multiplication value calculation unit 105 calculates a multiplication value by which the color signals RIN, GIN, and BIN are multiplied from the maximum value detected by the maximum value detection unit 103 and the tone correction table created by the tone correction table creation unit 102. Calculate
For example, the multiplication value calculation unit 105 calculates a multiplication value for the maximum value detected by the maximum value detection unit 103 from the gradation correction table in order to make the output gradation corresponding to the maximum value. . Here, by multiplying the maximum value, the multiplied value is calculated so that the output gradation corresponding to the maximum value is obtained.

FIG. 5 is a graph for explaining a method of calculating a multiplication value according to Embodiment 1. FIG.
FIG. 5 shows a polygonal line plotting the values shown in the gradation correction table.
Then, the multiplication value calculation unit 105 calculates the slope X of the point P corresponding to the maximum value CMAX detected by the maximum value detection unit 103 on the polygonal line.
For example, the multiplication value calculation unit 105 obtains the slope X by weighted average from the slope A and the slope B, which can be found from the output gradation of the break points on both sides of the maximum value CMAX. Specifically, as shown in FIG. 5, assuming that the maximum value CMAX is positioned 1 from the pixel value a and 4 from the pixel value b, the slope X can be asked for.
Inclination X = (Inclination A x 4 + Inclination B x 1)/5 (1)
Then, the multiplication value calculation section 105 gives the calculated slope X to the multiplication value correction section 106 as a multiplication value.

Multiplied value correcting section 106 calculates, if necessary, multiplied value calculating section 105 based on the maximum value detected by maximum value detecting section 103 and the minimum value detected by minimum value detecting section 104. modifies the multiplied value.
For example, when the multiplied value is 1 or more, the multiplied value correcting section 106 gives the multiplied value as a corrected multiplied value to the multipliers 107A, 107B, and 107C. On the other hand, when the multiplication value is less than 1, multiplied value correction section 106 determines that the difference between the maximum value detected by maximum value detection section 103 and the minimum value detected by minimum value detection section 104 is A corrected multiplied value is obtained by correcting the multiplied value so that it approaches 1 as the value increases. Specifically, the multiplication value correction unit 106 calculates a corrected multiplication value by correcting the multiplication value as follows. The modified multiplied values are provided to multipliers 107A, 107B, 107C.

Assuming that the gradation of the color signals RIN, GIN, and BIN is 10 bits, the maximum value that the maximum value CMAX detected by the maximum value detection unit 103 and the minimum value CMIN detected by the minimum value detection unit 104 can take is 1,023. Further, if an arbitrary set value for setting the strength of correction is G, the correction multiplication value can be obtained by the following equation (2).
Correction Multiplier =
1-(1-multiplication value)*(1-(CMAX-CMIN)*G/1023)
(2)

Here, if the set value G is greater than 1, the value of "1-(CMAX-CMIN)×G/1023" in equation (2) may become negative. In that case, by setting the value of this part to 0, the corrected multiplication value becomes 1.

As is clear from equation (2), when CMAX=CMIN, the corrected multiplication value is the multiplication value. In other words, when the gradations of each color are all the same in a pixel, the multiplication value is not modified. Note that when the gradation of each color is the same, the pixel becomes an achromatic color with no color.

On the other hand, when the difference between the maximum value CMAX and the minimum value CMIN is large, in other words, when the color of the pixel is deep and the saturation is high, the correction multiplication value becomes larger than the multiplication value and approaches 1. never exceed.

Multipliers 107A, 107B, and 107C respectively generate modified input image signals ROUT, GOUT, and BOUT by multiplying the color signals RIN, GIN, and BIN by the modified multiplied values output from the multiplied value modifying unit 106, A multiplication section outputs the modified input image signals ROUT, GOUT, and BOUT.

Some or all of the feature amount extraction unit 101, the tone correction table creation unit 102, the maximum value detection unit 103, the minimum value detection unit 104, the multiplication value calculation unit 105, and the multiplication value correction unit 106 described above may be, for example, , as shown in FIG. 6A, it can be composed of a memory 150 and a processor 151 such as a CPU (Central Processing Unit) that executes a program stored in the memory 150 . Such a program may be provided through a network, or recorded on a recording medium and provided. That is, such programs may be provided as program products, for example. In other words, the image processing apparatus 100 can be realized by a so-called computer.

In addition, part of the feature amount extraction unit 101, the tone correction table generation unit 102, the maximum value detection unit 103, the minimum value detection unit 104, the multiplication value calculation unit 105, the multiplication value correction unit 106, and the multipliers 107A, 107B, and 107C or all, for example, as shown in FIG. It can also be configured by a processing circuit 152 such as a programmable gate array.
As described above, the feature amount extraction unit 101, the tone correction table generation unit 102, the maximum value detection unit 103, the minimum value detection unit 104, the multiplication value calculation unit 105, the multiplication value correction unit 106, and the multipliers 107A, 107B, and 107C can be implemented by processing circuitry.

FIG. 7 is a flow chart showing the operation of the image processing apparatus 100 according to the first embodiment.
First, the image processing apparatus 100 receives an input image signal via an input terminal (not shown) functioning as an input unit (S10). The received input image signal is given to the feature extraction unit 101 , the maximum value detection unit 103 and the minimum value detection unit 104 .

The feature quantity extraction unit 101 extracts the feature quantity of the input image signal (S11). The extracted feature amount is given to the gradation correction table creation unit 102 .
The gradation correction table creation unit 102 creates a gradation correction table from the feature amount of the input image signal for one input image such as one frame (S12). The created gradation correction table is given to the multiplication value calculation unit 105 .

Further, the maximum value detection unit 103 detects the maximum value of gradation for each pixel of the input image indicated by the input image signal (S13). The detected maximum value is provided to multiplication value calculation section 105 and multiplication value correction section 106 .
The minimum value detection unit 104 detects the minimum value of gradation for each pixel of the input image indicated by the input image signal (S14). The detected minimum value is provided to multiplication value correction section 106 .

The multiplication value calculation unit 105 calculates a multiplication value from the gradation correction table and the maximum value (S15). The calculated multiplication value is provided to multiplication value correction section 106 .
The multiplied value correcting unit 106 specifies the corrected multiplied value by correcting the multiplied value using the maximum value and the minimum value as necessary (S16). The specified modified multiplication values are provided to multipliers 107A, 107B, 107C.

The multipliers 107A, 107B, and 107C multiply the input image signals by the modified multiplication values to perform gradation correction (S17). The modified input image signal, which is the input image signal modified by multiplying the modified multiplication value, is output from, for example, an output terminal (not shown) functioning as an output section.

If the image processing apparatus 100 does not particularly need to finish receiving the input image signal, for example, (Yes in S18), the process returns to step S10 and repeats the above process.
It should be noted that all the steps described above are generally repeated for time-sequential input of input image signals.

As described above, according to Embodiment 1, by multiplying the input image signal by the modified multiplication value, it is possible to suppress excessive reduction in the lightness of pixels with high saturation and vivid colors. As a result, power saving can be achieved without impairing the vividness of colors and visibility for the user.

Note that, for example, in order to facilitate digital calculation, the following formula (3) may be used instead of the above formula (2). Furthermore, the following formula (4) may be used.
Correction Multiplier =
1-(1-multiplication value)*(1-(CMAX-CMIN)*G/1024)
(3)
Correction Multiplier =
1-(1-multiplication value)*(1-(CMAX-CMIN)*G)
(4)

As in formula (2), in formula (3), if the part "1-(CMAX-CMIN)×G/1024" is negative, the value of this part is set to 0, and in formula (4), If the "1-(CMAX-CMIN)×G" portion is negative, the value of this portion is set to 0.

Embodiment 2.
FIG. 8 is a block diagram schematically showing the configuration of an image processing apparatus 200 according to Embodiment 2. As shown in FIG.
The image processing apparatus 200 includes a feature amount extraction unit 201, a tone correction table generation unit 102, a multiplication value calculation unit 205, a multiplication value correction unit 206, multipliers 107A, 107B, and 107C, and a table reference value identification unit. 208 and a maximum value identification unit 209 .

The gradation correction table creation unit 102 and the multipliers 107A, 107B, and 107C of the image processing device 200 according to the second embodiment are similar to the gradation correction table creation unit 102 and the multiplier 107A of the image processing device 100 according to the first embodiment. , 107B and 107C.

Note that FIG. 8 omits a delay unit for delaying the input image signal. The image processing apparatus 200 is assumed to include such a delay section to delay the input image signal.

The feature amount extraction unit 201 outputs a luminance signal indicating the luminance of the first color, and a color difference indicating the color difference between the first color and the second and third colors different from the first color. A predetermined feature quantity is extracted for each pixel from an input image signal composed of signals. In Embodiment 2, the feature amount extraction unit 201 extracts feature amounts from the luminance signal YIN and the color difference signals PbIN and PrIN as input image signals.
An input image signal in the second embodiment is composed of a luminance signal YIN and color difference signals PbIN and PrIN. The color difference signals PbIN and PrIN are 0 when the image signal has no color such as achromatic, white, black or gray. Also, the color difference signals PbIN and PrIN have positive or negative values when there is color, and the absolute value increases as the color becomes darker. The color difference signal may have an offset added to the value when there is no color, but for convenience of calculation, the value when there is no color is assumed to be 0 here.

A table reference value specifying unit 208 is a reference value specifying unit that specifies a reference value for referring to the tone correction table for each pixel from the input image signal. In the second embodiment, the table reference value specifying unit 208 extracts a reference value from the luminance signal YIN and the color difference signals PbIN and PrIN that form the input image signal. Identify reference values.
For example, the table reference value specifying unit 208 may calculate pixel values of R, G, and B from the luminance signal YIN and the color difference signals PbIN and PrIN, and use the maximum value as the table reference value for each pixel. Note that the table reference value identification unit 208 may use, for example, the luminance of each pixel indicated by the luminance signal YIN as the table reference value.
The table reference value is given to multiplication value calculation section 205 .

A multiplication value calculation unit 205 calculates a multiplication value for the table reference value from the gradation correction table in order to set the table reference value to the output gradation corresponding to the table reference value. Here, by multiplying the table reference value, the multiplication value is calculated so that the output gradation corresponding to the table reference value is obtained.
For example, the multiplication value calculation unit 205 calculates a multiplication value by which the input image signal is multiplied from the table reference value calculated by the table reference value identification unit 208 and the gradation correction table created by the gradation correction table creation unit 102. Calculate
The processing in multiplication value calculation section 205 in Embodiment 2 is the same as the processing in multiplication value calculation section 105 in Embodiment 1, except that a table reference value is used.

The maximum value specifying unit 209 specifies the maximum absolute value of the color difference for each pixel from the input image signal. For example, the maximum value specifying unit 209 specifies, for each pixel, the maximum value of the absolute values of the color differences indicated by the color difference signals PbIN and PrIN included in the input image signal. The identified maximum value is provided to multiplication value correction section 206 .

Multiplied value correction section 206 corrects the multiplied value calculated by multiplied value calculation section 205 based on the maximum value specified by maximum value specifying section 209, if necessary.
For example, when the multiplied value is 1 or more, the multiplied value correcting section 206 gives the multiplied value as a corrected multiplied value to the multipliers 107A, 107B, and 107C. On the other hand, when the multiplied value is less than 1, the multiplied value correction unit 206 corrects the multiplied value so that the larger the maximum value specified by the maximum value specifying unit 209 is, the closer to 1 the multiplied value is. . Specifically, multiplied value correction section 206 calculates corrected multiplied values by correcting multiplied values as follows, and supplies the corrected multiplied values to multipliers 107A, 107B, and 107C.

Assuming that the gradation of the input image signal is 10 bits, the maximum value that the maximum value PbPrMAX specified by the maximum value specifying unit 209 can take is 512. Further, if an arbitrary setting value for setting the strength of correction is G, the correction multiplication value can be obtained by the following equation (5).
Correction Multiplier =
1-(1-multiplication value) x (1-PbPrMAX x G/512) (5)

Here, if a value greater than 1 is set as the set value G, the value of "1-PbPrMAX×G/512" in the formula (5) may become negative. By setting the value of the part to 0, the modified multiplication value becomes 1. Further, as in the first embodiment, the relationship between the set value G and the division value 512 is arbitrary in the equation (5) as well.

As is clear from the equation (5), when PbPrMAX=0, the corrected multiplication value is the multiplication value. In other words, when the maximum color difference absolute value indicated by the color difference signals PbIN and PrIN is 0 in a certain pixel, the multiplication value is not modified. The maximum absolute value of the color difference is 0 when there is no color.
When the maximum absolute value of the color difference indicated by the color difference signals PbIN and PrIN is large, in other words, when the color is dark, the corrected multiplication value becomes larger than the multiplication value, approaches 1, and never exceeds 1. do not have.

As described above, when the multiplied value is 1 or more, the multiplied value correction unit 206 outputs the multiplied value as it is to the multipliers 107A, 107B, and 107C as the modified multiplied value, and when the multiplied value is less than 1, (5) Corrected multiplication values calculated using the equations are output to multipliers 107A, 107B, and 107C.

As described above, in Embodiment 2, by multiplying the input image signal by the modified multiplication value, it is possible to suppress excessive reduction in the lightness of pixels with high saturation and vivid colors. As a result, power saving can be achieved without impairing the vividness of colors and visibility for the user.

Embodiment 3.
FIG. 9 is a block diagram schematically showing the configuration of an image processing apparatus 300 according to Embodiment 3. As shown in FIG.
The image processing apparatus 300 includes a maximum value detection unit 103, a minimum value detection unit 104, a multiplication value calculation unit 305, a multiplication value correction unit 106, multipliers 107A, 107B, and 107C, and an area feature value calculation unit 310. , a feature amount synthesis unit 311 , a tone correction table storage unit 312 , and a tone correction table selection unit 313 .

Maximum value detection unit 103, minimum value detection unit 104, multiplied value correction unit 106, and multipliers 107A, 107B, and 107C of image processing apparatus 300 according to the third embodiment are the maximum value of image processing apparatus 100 according to the first embodiment. This is similar to the value detection section 103, the minimum value detection section 104, the multiplication value correction section 106, and the multipliers 107A, 107B, and 107C.

Although a delay unit for delaying the color signals RIN, GIN, and BIN is omitted in FIG. , the image processing apparatus 300 is provided with such a delay unit to delay the color signals RIN, GIN, and BIN.

The area feature amount calculation unit 310 calculates an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel. calculate.
In Embodiment 3, the area feature amount calculator 310 calculates area feature amounts from the color signals RIN, GIN, and BIN.
For example, the area feature amount calculation unit 310 extracts feature amounts such as brightness or brightness from the color signals RIN, GIN, and BIN, and calculates the average value of the feature amounts for each area that is a part of one image. , the area feature amount in each area is calculated. The calculated region feature amount is provided to the feature amount synthesizing unit 311 .

The feature quantity synthesizing unit 311 calculates a synthesized feature quantity by synthesizing the region feature quantity for each pixel.
For example, when one of a plurality of pixels included in an input image signal is set as a target pixel, the feature amount synthesizing unit 311 calculates the area feature of the target area including the target pixel. A combined feature amount is calculated by averaging the amount and the area feature amount of an area having a predetermined relationship with the target area after weighting according to the distance from the target pixel. Here, the predetermined relationship is an adjacency relationship, but it may be a relationship within a predetermined range with respect to the target area.
Specifically, the feature amount synthesizing unit 311 identifies a target pixel, which is a target pixel, from a plurality of pixels included in each region, and based on the positional relationship between the target pixel and each region, Synthesize region features. Then, the feature amount synthesizing section 311 provides the synthetic feature amount calculated for each pixel to the tone correction table selecting section 313 .

FIGS. 10A and 10B are schematic diagrams for explaining the process of calculating the synthetic feature amount.
For example, as shown in FIG. 10(A), it is assumed that area feature amounts are calculated for areas 11-13, areas 21-23, and areas 31-33.
Then, when calculating the composite feature amount for the pixel 22-1 in the area 22, as shown in FIG. A composite feature amount may be calculated by interpolation using a linear function, for example, according to the horizontal position x and vertical position y of the pixel 22-1. In this case, weighted averaging is performed by weighting according to the horizontal position x and the vertical position y. For example, the combined feature amount of the pixel a may be calculated by combining these area feature amounts so that the larger the horizontal position x and the vertical position y, the smaller the weighting.

Alternatively, the feature amount synthesizing unit 311 holds, as a table, functions of HWEIGHT(x) and VWEIGHT(y) corresponding to the horizontal position x and the vertical position y as shown in FIG. 5) The combined feature amount may be calculated by a formula in which the value determined by the function is normalized with a maximum value of 1.
(BL11×(1−HWEIGHT(x))+BL12×HWEIGHT(x))×
(1-VWEIGHT(y))+(BL21×(1-HWEIGHT(x))+
BL22×HWEIGHT(x))×VWEIGHT(y) (5)
Here, BL11 is the area feature amount of the area 11 shown in FIG. 10B, BL12 is the area feature amount of the area 12 shown in FIG. 10B, and BL21 is the area feature amount of FIG. ) and BL22 are the area feature amounts of the area 22 shown in FIG. 10B.

The gradation correction table storage unit 312 stores a plurality of gradations that indicate the correspondence relationship between the input gradation that is the gradation corresponding to the synthetic feature amount and each output gradation that is the gradation obtained by correcting the input gradation. A gradation correction information storage unit that stores correction information.
In Embodiment 3, the tone correction table storage unit 312 stores a plurality of tone correction tables. It is assumed that the plurality of gradation correction tables are configured such that the smaller the synthetic feature value, the brighter the correction, and the larger the synthetic feature value, the darker the correction.
In other words, the plurality of gradation correction tables differ in the brightness of the output gradation corresponding to the input gradation.

FIG. 12 is a schematic diagram showing a graph plotting a plurality of gradation correction tables stored in the gradation correction table storage unit 312. As shown in FIG.
FIG. 12 shows an example of plotting 17 gradation correction tables.
Assume that an identification number from 0 to 16 is assigned to each of the 17 gradation correction tables shown in FIG. Here, it is assumed that identification numbers 0 to 16 are assigned in order from the top of FIG.

A gradation correction table selection unit 313 performs gradation correction information selection that selects gradation correction information corresponding to the composite feature amount for each pixel from a plurality of gradation correction information stored in the gradation correction table storage unit 312 . Department.
For example, the gradation correction table selection unit 313 selects one of the plurality of gradation correction tables stored in the gradation correction table storage unit 312 using the synthetic feature amount of each pixel provided from the feature amount synthesis unit 311 . Select the gradation correction table to use. In Embodiment 3, the gradation correction table selection unit 313 selects a gradation correction table that associates a darker output gradation with an input gradation as the brightness of the synthetic feature amount becomes brighter.

For example, when the combined feature amount is 8 bits, the gradation correction table selection unit 313 selects a set of two gradation correction tables from the upper 4 bits of the combined feature amount, and with the lower 4 bits, Take a weighted average of the pairs.
Specifically, the gradation correction table selection unit 313 selects a gradation correction table whose identification numbers are 0 and 1 when the high-order 4 bits of the composite feature amount is 0, and when the high-order 4 bits are 1, the identification number is If the upper 4 bits are 15, then the gradation correction tables 15 and 16 with identification numbers 15 and 16 are selected.

Then, the gradation correction table selection unit 313 synthesizes two tables by weighted averaging for each of the breaking points obtained by the 16 divisions as described above.
Specifically, when the gradation correction tables with identification numbers 0 and 1 are selected, the gradation correction table selection unit 313 selects the gradation with the identification number 0 when the low-order 4 bits of the synthetic feature amount is 0. The correction table is multiplied by (16-0), the gradation correction table whose identification number is 1 is multiplied by 0, added together, and divided by 16. Further, when the low-order 4 bits of the composite feature amount are 1, the gradation correction table selection unit 313 multiplies the gradation correction table with the identification number of 0 by (16-1), and obtains the gradation correction table with the identification number of 1. Multiply the table by 1, add them together and divide by 16. Similarly, when the low-order 4 bits of the composite feature value is 15, the gradation correction table selection unit 313 multiplies the gradation correction table with the identification number of 0 by (16-15), and calculates the gradation with the identification number of 1. Multiply the correction table by 15, add them together and divide by 16.

In the above description, 17 gradation correction tables are stored in the gradation correction table storage unit 312, but the third embodiment is not limited to such an example.
For example, 256 tone correction tables may be stored in the tone correction table storage unit 312 . In such a case, the gradation correction table selection unit 313 may select a predetermined gradation correction table according to the composite feature amount. Specifically, the gradation correction table selection unit 313 selects a gradation correction table with an identification number of 0 when the combined feature amount is 0, and a gradation correction table with an identification number of 1 when the combined feature amount is 1. You can make a selection like so. Also in this case, as shown in FIG. 12, the lower the identification number of the gradation correction table, the higher the graph of the table.
By making the tone correction table stored in the tone correction table storage unit 312 convex downward, the output tone becomes smaller than the input tone, and the power consumption is reduced.

Returning to FIG. 9, the multiplication value calculation unit 305 calculates the maximum value detected by the maximum value detection unit 103 for each pixel from the gradation correction information selected for each pixel, and calculates the output gradation corresponding to the maximum value. , a multiplication value for the maximum value is calculated. Here, by multiplying the maximum value, the multiplied value is calculated so that the output gradation corresponding to the maximum value is obtained.
For example, the multiplication value calculation unit 305 uses the gradation correction table for each pixel provided from the gradation correction table selection unit 313 to calculate the multiplication value. In Embodiment 1, the multiplication value is calculated based on one gradation correction table for one image, but in Embodiment 3, a different gradation correction table is used for each pixel. Note that the process of calculating the multiplication value using the maximum value from the gradation correction table is the same as in the first embodiment.

Embodiment 4.
FIG. 13 is a block diagram schematically showing the configuration of an image processing device 400 according to the fourth embodiment.
The image processing apparatus 200 includes a multiplication value calculation unit 405, a multiplication value correction unit 206, multipliers 107A, 107B, and 107C, a table reference value identification unit 208, a maximum value identification unit 209, and an area feature amount calculation unit 410. , a feature amount synthesis unit 311 , a tone correction table storage unit 312 , and a tone correction table selection unit 313 .

Multipliers 107A, 107B, and 107C of image processing apparatus 400 according to the fourth embodiment are the same as multipliers 107A, 107B, and 107C of image processing apparatus 100 according to the first embodiment.
The multiplication value correction unit 206, the table reference value identification unit 208, and the maximum value identification unit 209 of the image processing apparatus 400 according to the fourth embodiment are similar to the multiplication value correction unit 206, the table reference It is the same as the value specifying unit 208 and the maximum value specifying unit 209 .
The feature amount synthesizing unit 311, tone correction table storage unit 312, and tone correction table selecting unit 313 of the image processing apparatus 400 according to the fourth embodiment are similar to the feature amount synthesizing unit 311 of the image processing apparatus 300 according to the third embodiment. , the gradation correction table storage unit 312 and the gradation correction table selection unit 313 .

Note that FIG. 13 omits a delay unit for delaying the input image signal. The image processing apparatus 400 is assumed to include such a delay section to delay the input image signal.

The region feature amount calculation unit 410 generates a luminance signal indicating the luminance of the first color, and a color difference signal indicating the color difference between the first color and the second and third colors different from the first color. A region feature amount, which is a predetermined feature amount, is calculated for each predetermined region from the input image signal consisting of In Embodiment 4, the area feature amount calculation unit 410 calculates area feature amounts from the input luminance signal YIN and the color difference signals PbIN and PrIN that form the input image signal.
For example, the area feature amount calculation unit 410 calculates the RBG gradation from the input luminance signal YIN and the color difference signals PbIN and PrIN, and calculates the area feature amount in the same manner as in the third embodiment.
The calculated area feature amount is provided to the tone correction table selection section 313 .

A multiplication value calculation unit 405 multiplies the table reference value for each pixel from the gradation correction table selected for each pixel in order to set the table reference value to the output gradation corresponding to the table reference value. calculate. Here, by multiplying the table reference value, the multiplication value is calculated so that the output gradation corresponding to the table reference value is obtained.
For example, the multiplication value calculation unit 405 uses the gradation correction table for each pixel provided from the gradation correction table selection unit 313 to calculate the multiplication value. In Embodiment 2, the multiplication value is calculated based on one tone correction table for one image, but in Embodiment 4, a different tone correction table is used for each pixel. Note that the process of calculating the multiplication value from the gradation correction table using the table reference value is the same as in the second embodiment.

The image processing apparatuses 100 and 300 according to Embodiments 1 and 3 described above can be used in an image display apparatus 140 as shown in FIG.
Image display device 140 includes image processing device 100 or 300 according to Embodiment 1 or 3 and self-luminous device 141 .

The image processing apparatuses 100 and 300 perform image processing on the color signals RIN, GIN, and BIN, and output modified input image signals ROUT, GOUT, and BOUT to the self-luminous device 141 . The self-luminous device 141 displays an image based on the corrected input image signals ROUT, GOUT, BOUT output from the image processing devices 100, 300. FIG.

15, a first image processing device 142 is provided in front of the image processing devices 100 and 300, and a second image processing device An image processing device 143 may be provided.
In this case, the first image processing device 142 and the second image processing device 143 are devices that perform processing different from that of the image processing devices 100 and 300 .

For example, the first image processing device 142 may be a device that generates the color signals RIN, GIN, and BIN by performing signal processing for removing noise from the original input image signals RIN#, GIN#, and BIN#. good.
The second image processing device 143 also performs signal processing for increasing the sharpness of the corrected input image signals ROUT, GOUT, BOUT to generate output image signals ROUT#, GOUT#, BOUT#.
In this case, self-luminous device 141 # displays an image based on output image signals ROUT#, GOUT#, and BOUT# output from second image processing device 143 .

14 and 15, the image processing apparatuses 100 and 300 according to the first or third embodiment are used, but the image processing apparatuses 100 and 300 according to the first or third embodiment are replaced with Image processing apparatuses 200 and 400 according to form 2 or 4 may be used.
In this case, the luminance signal YIN and the color difference signals PbIN and PrIN are input to the image processing apparatuses 200 and 400 as input image signals.

100, 200, 300, 400 image processing device, 101, 201 feature amount extraction unit, 102 gradation correction table creation unit, 103 maximum value detection unit, 104 minimum value detection unit, 105, 205, 305, 405 multiplication value calculation unit , 106, 206 multiplication value correction unit, 107 multiplier, 208 table reference value identification unit, 209 maximum value identification unit, 310, 410 area feature amount calculation unit, 311 feature amount synthesis unit, 312 gradation correction table storage unit, 313 Gradation correction table selector, 140, 140# image display device, 141, 141# self-luminous device, 142 first image processing device, 143 second image processing device.

Claims (18)

1. a feature quantity extraction unit for extracting a predetermined feature quantity for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel;
   The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. a gradation correction information creation unit that creates gradation correction information indicating a correspondence relationship between a gradation and an output gradation that is a gradation obtained by correcting the input gradation;
   a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal;
   a multiplication value calculation unit that calculates a multiplication value for the maximum value from the gradation correction information so as to make the maximum value the output gradation corresponding to the maximum value;
   a minimum value detection unit that detects the minimum value of the gradation for each pixel from the input image signal;
   When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the larger the difference between the maximum value and the minimum value, the more the value becomes 1. a multiplication value correcting unit that sets the corrected multiplication value to a value obtained by correcting the multiplication value so as to be closer to each other;
   and a multiplication unit that multiplies the input image signal by the modified multiplication value.
2. an area feature amount calculation unit that calculates an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel; ,
   a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel;
   Gradation correction that stores a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation. an information storage unit;
   a gradation correction information selection unit that selects gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information;
   a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal;
   a multiplication value calculation unit for calculating a multiplication value for the maximum value in order to set the maximum value to the output gradation corresponding to the maximum value for each pixel from the gradation correction information selected for each pixel; ,
   a minimum value detection unit that detects the minimum value of the gradation for each pixel from the input image signal;
   When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the larger the difference between the maximum value and the minimum value, the more the value becomes 1. a multiplication value correcting unit that sets the corrected multiplication value to a value obtained by correcting the multiplication value so as to be closer to each other;
   and a multiplication unit that multiplies the input image signal by the modified multiplication value.
3. When one of a plurality of pixels included in the input image signal is set as a target pixel, the feature amount synthesizing unit performs the area of the target area, which is the area including the target pixel. The feature amount and the area feature amount of the area having a predetermined relationship with respect to the target area are weighted according to the distance from the target pixel and averaged to obtain the combined feature. 3. The image processing apparatus according to claim 2, wherein the amount is calculated.
4. wherein the plurality of gradation correction information are different in brightness of the output gradation corresponding to the input gradation;
   2. The gradation correction information selection unit selects the gradation correction information that associates the input gradation with the output gradation that is brighter as the brightness of the synthetic feature amount becomes brighter. 3. The image processing apparatus according to 3 above.
5. When the multiplication value is less than 1, the multiplication value correction unit calculates the correction multiplication value using the following formula (1) with G as a predetermined set value. 1-(1- Multiplied value)×{1−(Maximum value−Minimum value)×G} (1)
   The image processing apparatus according to any one of claims 1 to 4, characterized by:
6. from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color , a feature quantity extraction unit for extracting a predetermined feature quantity for each pixel;
   The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. a gradation correction information creation unit that creates gradation correction information indicating a correspondence relationship between a gradation and an output gradation that is a gradation obtained by correcting the input gradation;
   a reference value identifying unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal;
   a multiplication value calculation unit that calculates a multiplication value for the reference value from the gradation correction information in order to set the reference value to the output gradation corresponding to the reference value;
   a maximum value identifying unit that identifies the maximum absolute value of the color difference for each pixel from the input image signal;
   When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the multiplication value is modified so as to approach 1 as the maximum value increases. a multiplication value correction unit that sets a value to the correction multiplication value;
   and a multiplication unit that multiplies the input image signal by the modified multiplication value.
7. from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color an area feature amount calculation unit that calculates an area feature amount, which is a predetermined feature amount, for each predetermined area;
   a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel;
   Gradation correction that stores a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation. an information storage unit;
   a gradation correction information selection unit that selects gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information;
   a reference value identifying unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal;
   a multiplication value calculation unit for calculating a multiplication value of the reference value from the gradation correction information selected for each pixel so as to set the reference value to the output gradation corresponding to the reference value for each pixel; ,
   a maximum value identifying unit that identifies the maximum absolute value of the color difference for each pixel from the input image signal;
   When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the multiplication value is modified so as to approach 1 as the maximum value increases. a multiplication value correction unit that sets a value to the correction multiplication value;
   and a multiplication unit that multiplies the input image signal by the modified multiplication value.
8. When one of a plurality of pixels included in the input image signal is set as a target pixel, the feature amount synthesizing unit performs the area of the target area, which is the area including the target pixel. The feature amount and the area feature amount of the area having a predetermined relationship with respect to the target area are weighted according to the distance from the target pixel and averaged to obtain the combined feature. The image processing apparatus according to claim 7, wherein the amount is calculated.
9. wherein the plurality of gradation correction information are different in brightness of the output gradation corresponding to the input gradation;
   8. The gradation correction information selection unit selects the gradation correction information that associates the input gradation with the output gradation that is brighter as the brightness of the synthetic feature amount becomes brighter. Or the image processing device according to 8.
10. When the multiplication value is less than 1, the multiplication value correction unit calculates the correction multiplication value using the following formula (2) with G as a predetermined set value. 1-(1- Said multiplied value)×{1−said maximum value×G} (2)
    The image processing apparatus according to any one of claims 6 to 9, characterized by:
11. the computer,
    A feature quantity extraction unit that extracts a predetermined feature quantity for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel;
    The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. a gradation correction information creation unit that creates gradation correction information indicating a correspondence relationship between a gradation and an output gradation that is a gradation obtained by correcting the input gradation;
    a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal;
    a multiplication value calculation unit that calculates a multiplication value for the maximum value from the gradation correction information in order to set the maximum value to the output gradation corresponding to the maximum value;
    a minimum value detection unit that detects the minimum value of the gradation for each pixel from the input image signal;
    When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the larger the difference between the maximum value and the minimum value, the more the value becomes 1. a multiplication value correction unit that sets the corrected multiplication value to a value obtained by correcting the multiplication value so that the multiplication value is closer to the multiplication value;
    A program functioning as a multiplier that multiplies the input image signal by the modified multiplication value.
12. the computer,
    an area feature amount calculation unit that calculates an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel;
    a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel;
    Gradation correction that stores a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation. information storage unit,
    a gradation correction information selection unit that selects gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information;
    a maximum value detection unit that detects the maximum value of the gradation for each pixel from the input image signal;
    a multiplication value calculation unit for calculating, for each pixel, a multiplication value for the maximum value from the gradation correction information selected for each pixel, in order to make the maximum value the output gradation corresponding to the maximum value;
    a minimum value detection unit that detects the minimum value of the gradation for each pixel from the input image signal;
    When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the larger the difference between the maximum value and the minimum value, the more the value becomes 1. a multiplication value correction unit that sets the corrected multiplication value to a value obtained by correcting the multiplication value so that the multiplication value is closer to the multiplication value;
    A program functioning as a multiplier that multiplies the input image signal by the modified multiplication value.
13. the computer,
    from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color , a feature quantity extraction unit for extracting a predetermined feature quantity for each pixel;
    The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. a gradation correction information creation unit that creates gradation correction information indicating a correspondence relationship between a gradation and an output gradation that is a gradation obtained by correcting the input gradation;
    a reference value identifying unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal;
    a multiplication value calculation unit that calculates a multiplication value for the reference value from the gradation correction information in order to make the reference value the output gradation corresponding to the reference value;
    a maximum value identifying unit that identifies a maximum absolute value of the color difference for each pixel from the input image signal;
    When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the multiplication value is modified so as to approach 1 as the maximum value increases. a multiplication value correction unit that sets a value to the correction multiplication value; and
    A program functioning as a multiplier that multiplies the input image signal by the modified multiplication value.
14. the computer,
    from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color , an area feature amount calculator that calculates an area feature amount, which is a predetermined feature amount, for each predetermined area;
    a feature amount synthesizing unit that calculates a synthesized feature amount by synthesizing the area feature amount for each pixel;
    Gradation correction that stores a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation. information storage unit,
    a gradation correction information selection unit that selects gradation correction information corresponding to the composite feature amount for each pixel from the plurality of gradation correction information;
    a reference value identifying unit that identifies a reference value for referring to the tone correction information for each pixel from the input image signal;
    a multiplication value calculation unit that calculates a multiplication value for the reference value for each pixel, from the gradation correction information selected for each pixel, in order to make the reference value the output gradation corresponding to the reference value;
    a maximum value identifying unit that identifies a maximum absolute value of the color difference for each pixel from the input image signal;
    When the multiplication value is 1 or more, the multiplication value is used as a modified multiplication value, and when the multiplication value is less than 1, the multiplication value is modified so as to approach 1 as the maximum value increases. a multiplication value correction unit that sets a value to the correction multiplication value; and
    A program functioning as a multiplier that multiplies the input image signal by the modified multiplication value.
15. Extracting a predetermined feature amount for each pixel from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel,
    The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. creating gradation correction information indicating a correspondence relationship between gradation and output gradation, which is a gradation obtained by correcting the input gradation;
    detecting the maximum value of the gradation for each pixel from the input image signal;
    calculating a multiplication value for the maximum value in order to set the maximum value to the output gradation corresponding to the maximum value from the gradation correction information;
    detecting the minimum value of the gradation for each pixel from the input image signal;
    when the multiplication value is 1 or more, the multiplication value is a modified multiplication value;
    when the multiplication value is less than 1, the modified multiplication value is a value obtained by correcting the multiplication value so as to approach 1 as the difference between the maximum value and the minimum value increases;
    An image processing method comprising: multiplying the input image signal by the modified multiplication value.
16. calculating an area feature amount, which is a predetermined feature amount, for each predetermined area from an input image signal composed of a plurality of color signals indicating the gradation of each color for each pixel;
    calculating a synthesized feature amount by synthesizing the region feature amount for each pixel;
    storing a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation;
    selecting tone correction information corresponding to the composite feature amount for each pixel from the plurality of tone correction information;
    detecting the maximum value of the gradation for each pixel from the input image signal;
    calculating, for each pixel, a multiplication value for the maximum value from the gradation correction information selected for each pixel, in order to make the maximum value the output gradation corresponding to the maximum value;
    detecting the minimum value of the gradation for each pixel from the input image signal;
    when the multiplication value is 1 or more, the multiplication value is a modified multiplication value;
    when the multiplication value is less than 1, the modified multiplication value is a value obtained by correcting the multiplication value so as to approach 1 as the difference between the maximum value and the minimum value increases;
    An image processing method comprising: multiplying the input image signal by the modified multiplication value.
17. from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color , extracting a predetermined feature amount for each pixel,
    The number of pixels is tallied for each division into which the feature amount is divided in ascending order, and the number of pixels is added in order of the division having the smallest feature amount for normalization, thereby inputting the gradation corresponding to the feature amount. creating gradation correction information indicating a correspondence relationship between gradation and output gradation, which is a gradation obtained by correcting the input gradation;
    identifying a reference value for referring to the tone correction information for each pixel from the input image signal;
    calculating a multiplication value for the reference value in order to set the reference value to the output gradation corresponding to the reference value from the gradation correction information;
    identifying the maximum absolute value of the color difference for each pixel from the input image signal;
    when the multiplication value is 1 or more, the multiplication value is a modified multiplication value;
    when the multiplication value is less than 1, the corrected multiplication value is a value obtained by correcting the multiplication value so as to approach 1 as the maximum value increases;
    An image processing method comprising: multiplying the input image signal by the modified multiplication value.
18. from an input image signal comprising a luminance signal indicating the luminance of a first color and a color difference signal indicating color differences between the first color and second and third colors different from the first color , calculating an area feature amount, which is a predetermined feature amount, for each predetermined area;
    calculating a synthesized feature amount by synthesizing the region feature amount for each pixel;
    storing a plurality of gradation correction information indicating a correspondence relationship between an input gradation that is a gradation corresponding to the composite feature amount and each output gradation that is a gradation obtained by correcting the input gradation;
    selecting tone correction information corresponding to the composite feature amount for each pixel from the plurality of tone correction information;
    identifying a reference value for referring to the tone correction information for each pixel from the input image signal;
    calculating, for each pixel, a multiplication value of the reference value from the gradation correction information selected for each pixel, in order to set the reference value to the output gradation corresponding to the reference value;
    identifying the maximum absolute value of the color difference for each pixel from the input image signal;
    when the multiplication value is 1 or more, the multiplication value is a modified multiplication value;
    when the multiplication value is less than 1, the corrected multiplication value is a value obtained by correcting the multiplication value so as to approach 1 as the maximum value increases;
    An image processing method comprising: multiplying the input image signal by the modified multiplication value.

Images