JP2007013666A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enhance an object and to make the object inconspicuous by adjusting contrast of a particular object in an image. <P>SOLUTION: An object histogram calculating part 9 extracts a first object 5 and a second object 6 included in an image 12, calculates luminance histograms 7 and 8, operates an object γ correction coefficient calculating part 11 to correct a γ curve on the basis of any one or more among distribution contents of the luminance histograms 7 and 8, average luminance, maximum luminance, and minimum luminance, and operates an object gradation correcting part 13 to apply γ correction processing to the first object 5 and the second object 6 on the basis of an object γ correction coefficient after calculating the object γ correction coefficient. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image processing apparatus that processes an image such as a television image, and more particularly to an image processing apparatus that improves the contrast of a specific image (object) included in a television image or the like.

In general, in an image processing apparatus that processes a television image or the like, first, as shown in FIG. 6A, the luminance of each pixel constituting the image 101 is extracted in order to improve the contrast of the entire image. The extracted pixels are histogrammed for each luminance value as shown in FIG. Based on the luminance histogram thus obtained, as shown in FIG. 6C, after correcting the γ curve simulating the characteristics of the human eye, an image is obtained using the γ correction coefficient obtained by the correction process. By uniformly correcting the entire image, the contrast of the entire image is improved.
JP 2000-350060 A JP 2003-50002 A

  However, in such an image processing apparatus, although the contrast of the image 101 can be increased or decreased as a whole, a specific image included in the image 101, for example, a human image 102, and a tree The image 103 or the like cannot be selected to increase or decrease the contrast. For this reason, development of an image processing apparatus capable of performing image processing of such a specific image has been strongly desired.

  SUMMARY OF THE INVENTION In view of the above circumstances, an object of the present invention is to provide an image processing apparatus capable of adjusting a contrast of a specific image specified in advance in an image to enhance the specific image or make it inconspicuous. Yes.

  In order to achieve the above object, the present invention provides an object extraction unit for extracting object information given in an incoming video signal, and an object histogram for calculating a luminance histogram of the object extracted by the object extraction unit. An object γ correction for performing γ correction based on one or more data of a calculation unit and luminance histogram distribution data, average luminance data, maximum luminance data, and minimum luminance data calculated by the object histogram calculation unit An object γ correction coefficient calculation unit that calculates a coefficient, and an object gradation correction unit that performs γ correction processing on the object based on the object γ correction coefficient calculated by the object γ correction coefficient calculation unit It is said.

  According to the image processing apparatus of the present invention, it is possible to adjust the contrast of a specific image (object) designated in advance in an image, thereby enhancing the specific image or making it inconspicuous.

  FIG. 1 is a block diagram showing an embodiment of an image processing apparatus according to the present invention.

  The image processing apparatus 1 shown in this figure takes in an image (video signal) such as a television image, extracts a specified specific image (hereinafter referred to as an object), and adjusts the contrast. A processing unit 2 and an entire image processing unit 3 that reconciles the contrast of the object processed by the specific image processing unit 2 with the contrast of the entire image are provided.

  The specific image processing unit 2 takes in object information instructed by an editor or the like, and outputs an object selection instruction 4 that outputs an object selection instruction indicating which part of the image is an object, and an image (video) such as a television image. Signal), the first object 5 and the second object 6 are extracted from the object designated by the object selection instruction output from the object extraction unit 4, for example, the image 12 shown in FIG. 5 and an object histogram calculation unit 9 for calculating the luminance histograms 7 and 8 by summing up the luminances of the respective pixels constituting the second object 6 separately for each size. Further, the specific image processing unit 2 takes in an image (video signal) such as a television image, temporarily stores it in a frame unit, and outputs it in a FIFO (First In First Out) system, and an object Based on the object selection instruction output from the extraction unit 4 and the luminance histograms 7 and 8 output from the object histogram calculation unit 9, it corresponds to the luminance distribution pattern, image occupancy, average luminance, maximum luminance, minimum luminance, and the like. As described above, the γ correction coefficient calculation for the first object 5 and the γ correction coefficient for the second object 6 are calculated by adjusting the γ curve for the first object 5 and the γ curve for the second object 6 respectively. Unit 11 and the object selection instruction and object γ output from the object extraction unit 4 Based on each γ correction coefficient output from the positive coefficient calculation unit 11, the first object 5 and the second object 6 are extracted from the image 12 output from the first frame memory unit 10, γ-corrected, and luminance And an object gradation correction unit 13 for correcting the above.

  Then, the specific image processing unit 2 captures the image 12 such as a television image based on the object information instructed by the editor or the like, and buffers the first image in the first frame memory unit 10 while specifying the first image 12 specified from the image 12. One object 5 and second object 6 are extracted, and luminance histograms 7 and 8 of these first object 5 and second object 6 are calculated. Further, based on these luminance histograms 7 and 8, after calculating γ correction coefficients corresponding to the luminance distribution pattern, image occupancy rate, average luminance, maximum luminance, minimum luminance, etc., based on these γ correction coefficients. The first object 5 and the second object 6 are extracted from the image 12 output from the first frame memory unit 10, the luminance is corrected based on each γ correction coefficient, and supplied to the entire image processing unit 3.

  The entire image processing unit 3 takes in an image output from the object gradation correction unit 13, temporarily stores it in units of frames, and outputs it in a FIFO (First In First Out) system, and an object gradation An image output from the correction unit 13 is captured, and an entire screen histogram calculation unit 19 that calculates a luminance histogram obtained by dividing the luminance of each pixel constituting the entire screen according to size and output from the entire screen histogram calculation unit 19. The entire screen that calculates the γ correction coefficient necessary to increase the slope of the γ curve near the luminance level with a large distribution and reduce the slope of the γ curve near the luminance level with a small distribution and a γ correction coefficient calculation unit 20.

  The entire image processing unit 3 uses the entire screen γ correction coefficient cyclic leak type integration unit 21 and the entire screen γ correction coefficient cyclic leak type integration unit 21 to output the second frame memory. The entire image output from the unit 18 is γ-corrected, and the entire screen tone correction unit 22 for adjusting the contrast of the entire screen is provided.

  As shown in FIG. 5, the entire screen γ correction coefficient cyclic leak type integrator 21 multiplies the current entire screen γ correction coefficient output from the entire screen γ correction coefficient calculation unit 20 by a cyclic coefficient smaller than “1”. An addition circuit 23 for adding the previous entire screen γ correction coefficient, and temporarily storing the entire screen γ correction coefficient output from the addition circuit 23 and holding the entire screen γ correction coefficient output as the previous entire screen γ correction coefficient Multiplying the circuit 24 by the previous entire screen γ correction coefficient output from the entire screen γ correction coefficient holding circuit 24 and a preset cyclic coefficient (a value smaller than “1”) and supplying the result to the adder circuit 23 Multiplying the multiplication circuit 25 to the previous screen overall γ correction coefficient holding circuit 24 by the gain correction coefficient set by the editor or the like, and the magnitude of the overall screen γ correction coefficient Adjust A multiplication circuit 26 that performs a cyclic leak type integration on the entire screen γ correction coefficient output from the entire screen γ correction coefficient calculation unit 20 to gently change the γ correction coefficient of the entire screen in the time direction. Even if a scene change or a sudden luminance change occurs, the discontinuous point of the γ correction coefficient in the time direction is prevented from occurring, and the first object 5 generated by the object unit γ correction processing in the specific image processing unit 2 Discontinuous points (discontinuous points in the time direction) are prevented from occurring in the second object 6 and the peripheral portions of the first object 5 and the second object 6.

  For example, as shown in FIG. 2, “human image” is designated as the first object 5 and “tree image” is designated as the second object 6 by the object information instructed by the editor or the like. It shall be. At this time, when the image 12 is supplied and buffered in the first frame memory unit 10, the “human image” and the “tree image” included in the image 12 are extracted, and each luminance intensity is extracted. Further, the number of pixels constituting the “human image” and the “tree image” is aggregated to calculate a luminance histogram. Thereafter, the γ slope of the luminance intensity portion with a large number of pixels (luminance level portion with a large distribution) is increased, and the γ gradient of the luminance intensity portion with a small number of pixels (luminance level portion with a small distribution) is decreased. The γ curves for the first object 5 and the second object 6 are corrected, and the luminance distribution of the luminance level portion with a large distribution is expanded, and the luminance distribution of the luminance level portion with a small distribution is reduced.

  In parallel with this operation, it is checked by an editor or the like whether or not an instruction is given to emphasize only a specific object to increase or decrease contrast. In this case, if an instruction to enhance or weaken the contrast by emphasizing only a specific object is set, regardless of the shape of the luminance histogram, the image occupation ratio, etc., the specified change rate will be The γ curve slope is changed. If an instruction for emphasizing only a specific object is not set, the corrected γ is set according to the shape of the luminance histogram of the first object 5 and the second object 6 with respect to the number of pixels of the entire image, the image occupation ratio, and the like. The slope of the curve is corrected.

  For example, if a “human image” is specified as a specific object to be emphasized and two “human images” are included in the image 12 as shown in FIG. "Is designated as the first object 14 and the second object 15, and the shapes of the luminance histograms 16 and 17 are checked. When the histogram shapes are the same, the image occupancy rates of the first object 14 and the second object 15, for example, the image occupancy rates of the first object 14 and the second object 15 are “30%” and “20”, respectively. % ”, When the slope of the γ curve with respect to the second object 15 changes by“ 1 ”according to the ratio“ 30/20 = 1.5 ”of each image occupancy rate, the γ curve with respect to the first object 14 The inclination is greatly changed by “1.5 times”.

  In parallel with this operation, the average brightness of the first object 5 and the second object 6 and the weighting coefficient shown in FIG. 4 are used to adjust the slope of the γ curve with respect to an object having a high average brightness. In addition, the slope of the γ curve for an object with a low average luminance is adjusted to be large, and the above-described image processing increases the γ slope of the luminance intensity portion with a large number of pixels (the luminance level portion with a large distribution). In addition, when the γ slope of the luminance intensity part with a small number of pixels (brightness level part with low distribution) is reduced, the high luminance level should not be too bright and the low luminance level should not be too dark. The slope change rate of the γ curve for the first object 5 and the second object 6 is adjusted.

  Furthermore, the upper limit value and the lower limit value for the slope change rate of the γ curve for the first object 5 and the second object 6 are adjusted according to the maximum brightness and the minimum brightness of the first object 5 and the second object 6, and the brightness After the movement amount upper limit value and the movement amount lower limit value for offsetting the entire histograms 7 and 8 are adjusted and the γ correction coefficients for the first object 5 and the second object 6 are calculated, based on these γ correction coefficients. The brightness of the first object 5 and the second object 6 is corrected, and an image in which the contrast of the first object 5 and the second object 6 is optimized is supplied to the overall processing unit 3.

  Then, in the entire image processing unit 3, the image output from the object gradation correction unit 13 is captured and buffered in the second frame memory unit 18. Is done. Thereafter, based on this luminance histogram, the entire screen γ correction coefficient calculation unit 20 increases the slope of the γ curve in the vicinity of the luminance level with a large distribution, and reduces the slope of the γ curve in the vicinity of the luminance level with a small distribution. A necessary entire screen γ correction coefficient is calculated. Next, the entire screen γ correction coefficient cyclic leak type integration unit 21 performs cyclic leak type integration to gently change the entire screen γ correction coefficient with respect to the time direction. Next, the entire screen gradation correction unit 22 corrects the luminance of the entire image output from the second frame memory unit 18 based on the entire screen γ correction coefficient, and performs the γ correction processing for each object in the specific image processing unit 2. The discontinuous points (discontinuous points in the time direction) are prevented from occurring in the first object 5 and the second object 6 generated in step 1 and the peripheral portions of the first object 5 and the second object 6. In addition, the contrast of the first object 5 and the second object 6 processed by the specific image processing unit 2 and the contrast of the entire image are harmonized so that even if a scene change or a sudden luminance change occurs, γ correction in the time direction Avoid discontinuities in coefficients.

  As described above, in this embodiment, the object histogram calculation unit 9 extracts the first object 5 and the second object 6 included in the image 12 to obtain the luminance histograms 7 and 8. Based on any one or more of the distribution content, average luminance, maximum luminance, and minimum luminance, the object γ correction coefficient calculation unit 11 is operated to correct the γ curve and calculate the object γ correction coefficient, and then the object γ Based on the correction coefficient, the object gradation correction unit 13 is operated to perform the γ correction processing on the first object 5 and the second object 6. For this reason, the contrast of the 1st object 5 and the 2nd object 6 in the image 12 can be adjusted, and a specific image can be emphasized or made inconspicuous.

  In this embodiment, when the object γ correction coefficient is calculated by the object γ correction coefficient calculation unit 11, the slope of the γ curve is based on the exclusive area ratio of each object, for example, the first object 14 and the second object 15. The rate of change is adjusted. For this reason, the exclusive area is large, and the contrast of the conspicuous first object 14 can be emphasized and adjusted.

  In this embodiment, the object gradation correction unit 13 performs gradation correction on each object, for example, the first object 5 and the second object 6, using the optimum object γ correction coefficient for each object. I try to do it. Therefore, the contrast of the first object 5 and the contrast of the second object 6 can be optimized according to the brightness histogram 7 of the first object 5 and the brightness histogram 8 of the second object 6.

  In this embodiment, the image in which the contrast of the first object 5 and the second object 6 is adjusted by the specific image processing unit 2 is guided to the entire image processing unit 3, and a luminance histogram of the entire image is calculated to calculate the γ curve. And a cyclic leak type integration is performed so that the change in the γ correction coefficient of the entire screen in the time direction becomes gentle. For this reason, discontinuous points (in the temporal direction) are not formed in the first object 5 and the second object 6 generated by the γ correction processing in the specific image processing unit 2 and the peripheral portions of the first object 5 and the second object 6. The contrast of the first object 5 and the second object 6 processed by the specific image processing unit 2 and the contrast of the entire image can be harmonized with each other, and scene changes and sudden brightness can be prevented. Even if a change occurs, it is possible to prevent a discontinuous point from occurring in the entire screen γ correction coefficient in the time direction.

  In the above-described embodiment, the first frame memory unit 10 and the second frame memory unit 18 buffer the image 12 and the like, and the processing delay of the object histogram calculation unit 9 and the object γ correction coefficient calculation unit 11 or The entire screen histogram calculation unit 9, the entire screen γ coefficient calculation unit 20, and the entire screen γ correction coefficient cyclic leak type integration unit 21 absorb the processing delay and adjust the phase difference. The first frame memory unit 10 and the second frame memory unit 18 may be omitted.

1 is a block diagram showing an embodiment of an image processing apparatus according to the present invention. It is a schematic diagram which shows the operation example of the image processing apparatus shown in FIG. It is a schematic diagram which shows the operation example of the object (gamma) correction coefficient calculation part shown in FIG. It is a schematic diagram which shows the operation example of the object (gamma) correction coefficient calculation part shown in FIG. FIG. 2 is a block diagram illustrating a detailed circuit configuration example of an entire screen γ correction coefficient cyclic leak type integration unit illustrated in FIG. 1. It is a schematic diagram which shows the operation example of the image processing apparatus known conventionally.

Explanation of symbols

1: Image processing device 2: Specific image processing unit 3: Whole image processing unit 4: Object extraction unit 5, 14: First object 6, 15: Second object 7, 8, 16, 17: Luminance histogram 9: Object histogram Calculation unit 10: First frame memory unit 11: Object γ correction coefficient calculation unit 12: Image 13: Object gradation correction unit 18: Second frame memory unit 19: Whole screen histogram calculation unit 20: Whole screen γ correction coefficient calculation unit 21: Whole screen γ correction coefficient cyclic leak type integration unit 22: Whole screen gradation correction unit 23: Addition circuit 24: Whole screen γ correction coefficient holding circuit 25, 26: Multiplication circuit

Claims (1)

An object extraction unit for extracting object information given in the incoming video signal;
An object histogram calculation unit for calculating a luminance histogram of the object extracted by the object extraction unit;
An object γ correction coefficient for performing γ correction based on any one or more of the distribution data, average luminance data, maximum luminance data, and minimum luminance data of the luminance histogram calculated by the object histogram calculation unit is calculated. An object γ correction coefficient calculation unit;
An object gradation correction unit that performs γ correction processing on the object based on the object γ correction coefficient calculated by the object γ correction coefficient calculation unit;
An image processing apparatus comprising:

Images