JPH09321988A - Gradation reduction device and gradation reduction method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the image quality of a display image by gradation reduction processing. SOLUTION: Multi-value image data are restored to a state of data not subject to gamma correction processing by an inverse gamma correction circuit 2 and stored in a memory 3. A gradation reduction circuit 6 uses at first a reference color table 4 to apply gradation reduction processing to multi-gradation image data stored in the memory 3. An error detection circuit 8 detects a gradation reduction error of the gradation reduction image data to apply it to a control circuit 9. The control circuit 9 compares the gradation reduction error with a prescribed threshold level and controls each section so that a reference color table is changed till the gradation reduction error reaches a prescribed threshold level or below, the gradation reduction processing is conducted again to detect the gradation reduction error when the error is the threshold level or over. Then the reference color able is used for a confirmed color table when the gradation reduction error reaches a prescribed threshold level or below and stored in the color table memory 5 and the gradation reduction processing is conducted for the image by using the confirmed color table.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a gradation reducing device and a gradation reducing method suitable for use in, for example, an image communication system, a monitor device, etc. The present invention relates to a gradation reduction device and a gradation reduction method for improving the image quality of a display image by performing gradation reduction processing in consideration of characteristics and / or display characteristics.

[0002]

2. Description of the Related Art Conventionally, for example, red image data (R),
The so-called full-color image data of 24 bits (16 million colors) each having 8 bits of green image data (G) and 8 bits of blue image data (B), RGB has 3 bits, 3 bits, and 2 bits, respectively. There is known a gradation reduction process for converting the image data into a total of 8-bit image data to reduce the recording rate, the transmission rate, and the like.

This gradation reduction processing is broadly divided as shown in FIG. 7 to form an optimum color table for each image.
Step (STEP1) and the second step (STEP2) of reducing the gradation only to the color that each color table has for each pixel
It is composed of

Specifically, such gradation reduction processing is
The image quality deterioration due to the gradation reduction process is evaluated by the mean square error in the color space such as RGB or L ^* a ^* b ^* by the so-called least square method, and the mean square error is minimized. Then, the color table is formed by using vector quantization (first step). Then, for each image data of each pixel, gradation is reduced by approximating with the color in the color table where the distance in the color space is the minimum (second step).

[0005]

However, the conventional gradation reduction apparatus simply evaluates image quality deterioration of an image based on the mean square error in a color space such as RGB or L ^* a ^* b ^*, and performs gradation reduction processing. Therefore, there is a problem that the visual deterioration of the image subjected to the gradation reduction processing (gradation reduced image) cannot be accurately evaluated, and a preferable display image cannot be surely obtained.

The present invention has been made in view of the above problems, and provides a gradation reducing apparatus and a gradation reducing method capable of performing a gradation reducing process for improving the image quality of a display image. To aim.

[0007]

In order to solve the above-mentioned problems, a gradation reduction apparatus according to the present invention performs gradation reduction processing on multi-gradation image information supplied from the outside to perform gradation reduction image information. From the error detecting means, and an error detecting means for detecting a gradation reducing error of the gradation reducing image information formed by the gradation reducing means based on human visual characteristics. And a control means for controlling the gradation reduction means so that the gradation reduction processing is performed on the multi-gradation image information of the image by the color table selected according to the gradation reduction error.

Such a gradation reduction device, when detecting the gradation reduction error of the gradation reduction image information formed by the gradation reduction means, carries out the detection based on human visual characteristics, and The gradation reduction processing is performed on the multi-gradation image information of the image by the color table having the minimum gradation reduction error.

Specifically, the control means changes the reference color table serving as a reference for the gradation reduction processing, and controls the gradation reduction means so that the gradation reduction processing is performed for each reference color table. At the same time, the gradation reduction error for each reference color table is detected by the error detection output from the error detection means. Then, the reference color table showing the minimum value of the gradation reduction error is detected as the color table for the gradation reduction processing of the image.

As a method of detecting the color table showing the minimum gradation reduction error, the reference color table is changed, and the gradation reduction error detected each time is compared with a predetermined threshold value. A reference color table having a gradation reduction error equal to or smaller than a predetermined threshold may be detected as a reference color table having a minimum gradation reduction error, or the reference color table is changed a predetermined number of times. The reference color table having the minimum gradation reduction error may be detected from the gradation reduction errors detected each time.

With this, the gradation reduction processing can be performed using the color table detected based on the human visual characteristics, so that the display image can be made visually excellent and the display image The image quality can be improved.

Next, the gradation reducing apparatus according to the present invention has, in addition to the above-mentioned structure, a structure having an inverse gamma correction means for performing an inverse gamma correction process on arbitrary multi-gradation image information supplied from the outside. Has become.

Thus, it is possible to restore the arbitrary multi-gradation image information which has been subjected to the gamma correction processing and which is supplied from the outside to the state in which the gamma correction processing has not been performed, and then perform the favorable gradation reduction processing. it can. Therefore, it is possible to provide an image communication service with higher image quality even in an image communication system in which display devices having various gradation display capabilities are mixed.

Next, in order to solve the above-mentioned problems, the tone reducing apparatus according to the present invention provides a color table used in the tone reducing process and arbitrary multi-tone image information supplied from the outside. A gradation reduction means for performing gradation reduction processing based on the color table to form gradation reduced image information, and changing the gradation reduced image information within the range of the color information of the color table, and A gradation reduction image information selecting unit that detects a gradation reduction error of each gradation reduction image information based on visual characteristics and selects and outputs the gradation reduction image information showing the minimum gradation reduction error. To do.

Such a gradation reduction device changes the gradation reduced image information formed by performing the gradation reduction processing using the reference color table within the range of the color information of the color table. At the same time, the gradation reduction error of each gradation reduction image information is detected based on the human visual characteristics. And
The gradation reduction image information indicating the minimum gradation reduction error is selected and output as the gradation reduction image information of the image.

Specifically, the color table that serves as a reference when performing the gradation reduction process is detected by the following means. That is, the gradation reduction device performs gradation reduction processing on the color table used in the gradation reduction processing and arbitrary multi-gradation image information supplied from outside based on the color table to perform gradation reduction. A gradation reduction unit that forms image information and an error detection unit that detects a gradation reduction error of the gradation reduction image information for each reference color table that has been subjected to the gradation reduction process by the color table variable gradation reduction unit. And have. Then, the gradation reducing means detects a reference color table showing the minimum value of the gradation reducing error detected by the error detecting means from each of the changed reference color tables, It is detected as a color table for performing the gradation reduction processing of the multi-gradation image information.

Further, the gradation reduced image information selecting means is
The gradation reduction image data is selected by the following means. That is, the gradation-reduced image information selection unit determines each pixel value of the gradation-reduced image information that has been subjected to gradation reduction processing based on the color table by the gradation reduction unit to the pixel value Pixel value changing means for changing within a range, and error detecting means for detecting a gradation reduction error of gradation reduction image information formed by gradation reduction processing based on the color table based on human visual characteristics. A tone reduction error of tone reduction image information formed by tone reduction processing based on the color table, and a tone reduction error of each tone reduction image information in which each pixel value is changed by the pixel value changing means. A step of controlling the error detecting means so as to detect each of them and selecting and outputting gradation-reduced image information showing the minimum gradation-reducing error among the gradation-reducing errors detected by the error detecting means. Key In the reduced image information output means, to select the grayscale reduced image data.

As a method of detecting the gradation-reduced image information showing the minimum gradation-reduction error, the gradation-reduced image information formed first is changed and the gradation-reduction error detected each time is changed. Alternatively, the gradation reduction image information having a gradation reduction error equal to or smaller than the predetermined threshold may be detected as the gradation reduction image information having the minimum gradation reduction error by comparing with a predetermined threshold. Alternatively, the gradation reduction image information formed at the beginning is changed a predetermined number of times, and the gradation reduction image information having the minimum gradation reduction error is selected from the gradation reduction errors detected each time. You may do it.

As a result, the gradation-reduced image information having the minimum gradation-reduction error detected based on human visual characteristics is selected from the gradation-reduced image information formed by the gradation-reduction processing. Since the image can be output, the displayed image can be made visually excellent, and the image quality of the displayed image can be improved.

Next, in the gradation reduction method according to the present invention, in order to solve the above-mentioned problems, gradation reduction processing is performed on multi-tone image information supplied from the outside to form gradation reduced image information. The gradation reduction error of the gradation reduction image information formed by the gradation reduction means is detected based on human visual characteristics, and a color table corresponding to the detected gradation reduction error is selected to select this color. The table is used to perform gradation reduction processing on the multi-gradation image information of the image.

Specifically, when the color table is selected, the reference color table including the color information serving as a reference when performing the gradation reduction processing of the image is changed, and the reference color table is changed for each reference color table. Applying gradation reduction processing to multi-gradation image information,
The gradation reduction error for each reference color table is detected. Then, the reference color table showing the minimum value of the gradation reduction error is selected as the color table.

Thus, the gradation reduction processing can be performed using the color table detected based on the human visual characteristics, so that the displayed image can be made visually excellent and the display image The image quality can be improved.

Next, the gradation reduction method according to the present invention solves the above-mentioned problems by gradation reduction processing of multi-gradation image information supplied from the outside by a color table used in the gradation reduction processing. The gradation reduction image information is formed by changing the gradation reduction image information within the range of the color information of the color table, and the gradation reduction error of each gradation reduction image information is corrected based on the human visual characteristics. The detected gradation reduction image information showing the minimum gradation reduction error is selected and output.

At this time, the color table used in the gradation reduction processing is the gradation reduction processing of the multi-gradation image information supplied from the outside while changing the reference color table including the reference color information. , Detecting the gradation reduction error of the gradation reduction image information for each of the reference color tables, and detecting the reference color table showing the minimum value of the gradation reduction error from among the changed reference color tables. Select by

The selection of the gradation reduced image information is performed by selecting each pixel value of the gradation reduced image information formed by the gradation reduction processing based on the color table within the range of the color information of the color table. While changing, the gradation reduction error of each gradation reduction image information is detected based on human visual characteristics, and the gradation showing the minimum gradation reduction error among the detected gradation reduction errors is displayed. This is done by selecting the reduced image information.

As a result, the gradation-reduced image information having the minimum gradation-reduction error detected based on the human visual characteristics is selected from the gradation-reduced image information formed by the gradation-reduction processing. Since the image can be output, the displayed image can be made visually excellent, and the image quality of the displayed image can be improved.

Next, in the gradation reduction method according to the present invention, in order to solve the above-mentioned problems, gradation reduction processing is performed on multi-tone image information supplied from the outside to form gradation reduced image information. , Detecting the gradation reduction error of the gradation reduction image information based on human visual characteristics, selecting a color table according to the gradation reduction error, and reducing the gradation of the multi-gradation image information of the image. The gradation-reduced image information that has been processed and gradation-reduced by the color table is changed within the range of the color information in the color table, and the gradation-reduction of each gradation-reduced image information is performed based on human visual characteristics. The error is detected, and the gradation-reduced image information indicating the minimum gradation-reduction error is selected and output.

Thus, the color table detected based on the human visual characteristics can be used for the gradation reduction processing, and the human being can be selected from the gradation reduction image information subjected to the gradation reduction processing using this color table. Since it is possible to select and output the best gradation-reduced image information detected based on the visual characteristics, the displayed image can be made even better visually, and the image quality of the displayed image can be further improved. Can be planned.

[0029]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of a gradation reducing apparatus and a gradation reducing method according to the present invention will be described in detail below with reference to the drawings. The gradation reducing apparatus and the gradation reducing method according to the present invention can be applied to the gradation reducing apparatus as shown in FIG. The gradation reducing apparatus according to the first embodiment of the present invention has, for example, a red image data (R), a green image data (G), and a blue image data (B) each having 8 bits each. 24-bit (1
An input terminal 1 to which multi-gradation image data of 6 million colors) is supplied, and an inverse gamma correction circuit 2 that performs inverse gamma correction processing on the multi-gradation image data that has already been subjected to gamma correction processing, A memory 3 for temporarily storing the multi-gradation image data and the like that has been subjected to the inverse gamma correction process, and for converting each RGB image data into 8-bit image data of 3 bits, 3 bits and 2 bits respectively. It has a reference color table 4 in which each color data serving as a reference is stored.

Further, the gradation reducing apparatus according to the first embodiment uses a color table memory 5 for storing a fixed color table (fixed color table) which will be described later, and the fixed color table and the like. A gradation reduction circuit 6 for performing gradation reduction processing of the multi-gradation image data into gradation reduction image data of 8 bits in total of 3 bits, 3 bits and 2 bits for RGB respectively.
And a visual characteristic ROM 7 in which the frequency characteristic of each pixel is stored as visual characteristic data, and an error for detecting an error in the gradation reduced image subjected to the gradation reduction processing based on the visual characteristic data from the visual characteristic ROM 7. It has a detection circuit 8.

Further, the gradation reducing apparatus according to the first embodiment has a control circuit 9 for controlling the entire apparatus and gradation-reduced image data subjected to gradation reduction processing by the gradation reducing circuit 6. In addition, a gamma correction circuit 10 that performs gamma correction processing according to the gamma characteristic of a display device that displays an image
And an output terminal 11 for outputting the gradation-reduced image data subjected to the gamma correction process.

Next, the operation of the gradation reducing device having such a configuration will be described with reference to the flowchart of FIG.
First, this flowchart starts when the main power source of the gradation reducing apparatus is turned on, and the process proceeds to step S1.

In step S1, the control circuit 9 detects the presence / absence of multi-gradation image data supplied to the inverse gamma correction circuit 2 to determine whether the image data is supplied to the gradation reduction device. If the determination is No, the step S1 is repeated until the image data is supplied, and Yes
In the case of s, it progresses to step S2.

In step S2, the inverse gamma correction circuit 2
However, the gamma correction process has already been performed, and the inverse gamma correction process is performed on the multi-gradation image data supplied through the input terminal 1 to perform the gamma correction process on the multi-gradation image data. It returns to the non-existing state and proceeds to step S3. After that, in the gradation reduction device, each processing such as gradation reduction proceeds on the multi-gradation image data in the state where the gamma correction processing is not performed.

Next, in step S3, the gradation reduction circuit 6
However, the gradation reduction processing of the multi-gradation image data is performed using the reference color table 4, and the process proceeds to step S4.

Specifically, when the multi-gradation image data is supplied from the inverse gamma correction circuit 2 to the control circuit 9,
This is once stored in the memory 3 as an original image for gradation reduction processing and read out, and the reference color table data is read from the reference color table 4 and supplied to the gradation reduction circuit 6. The gradation reduction circuit 6 first performs gradation reduction processing of the multi-gradation image data read from the memory 3 based on the reference color table data from the reference color table 4, and the error reduction circuit Supply to 8.

The human visual sensitivity is low at high frequencies where the brightness changes rapidly, but is low at low frequencies where the brightness changes gradually. It has the feature of high sensitivity. The visual characteristic ROM 7 stores visual characteristic data for each pixel formed based on such ergonomics.

Therefore, in step S4, the error detection circuit 8 detects the error of the image subjected to the gradation reduction processing by the reference color table data based on the visual characteristic data stored in the visual characteristic ROM 7. Step S5
Proceed to.

Specifically, the error detection circuit 8 forms a suitable sum of squares based on the observed value (gradation reduced image data in this case), and detects a value that minimizes it. The least squares method is used to detect the error (mean square error) of the image whose gradation has been reduced. At this time, the error detection circuit 8 uses the visual characteristic data as a weighting coefficient. As a result, the error of the gradation-reduced image detected by the error detection circuit 8 can be made to reflect the human visual characteristics. The mean square error detected by the error detection circuit 8 is temporarily stored in the memory in the control circuit 9.

Here, in the gradation reducing apparatus, as described below, the reference color table is changed a predetermined number of times, the above-mentioned mean square error is detected each time, and the mean square error is equal to or less than a predetermined threshold value. In this case, the reference color table in the case of is used as the fixed color table to perform the gradation reduction processing of the image. Therefore, in step S5,
The control circuit 9 determines whether or not the mean square error detected in step S4 is equal to or smaller than a predetermined threshold value. If No, the process proceeds to step S8, and if Yes, step S8.
Proceed to 6.

In step S8, the control circuit 9 minutely changes the reference color table 4 using a vector quantization codebook creation algorithm such as an algorithm such as Linde (LBG algorithm), and then proceeds to step S3. Return. Next, in step S5, N
If it is judged to be o, the reference color table is changed in step S8 and the process returns to step S3, the gradation reduction circuit 6
This time, based on this changed reference color table,
The gradation reduction processing of the multi-gradation image data stored in the memory 3 is performed, and the process proceeds to step S4. Then, in step S4, the error detection circuit 8 detects the mean squared error weighted by the visual characteristic, thereby detecting the error of the image in which the gradation is reduced based on the changed reference color table. Then, the mean square error is supplied to the control circuit 9 and the process proceeds to step S5. Step S above
In 5, the control circuit 9 determines whether or not the mean squared error of the changed reference color table is less than or equal to a predetermined threshold value. If it is less than or equal to the predetermined threshold value, the process proceeds to step S6. If it is equal to or more than the predetermined threshold, the above step S
In step 8, the reference color table is further changed and the process returns to step S3.

As described above, in the gradation reducing apparatus, the processes of steps S3 to S5 and step S8 are repeated until the mean square error becomes equal to or less than the predetermined threshold value. Then, when the mean square error becomes equal to or smaller than the predetermined threshold value, the control circuit 9 determines “Yes” in step S5 and proceeds to step S6.

Next, in step S6, the control circuit 9 is
However, as a result of minutely changing the reference color table, a reference color table showing a mean square error equal to or less than a predetermined threshold value is determined as a color table for performing gradation reduction processing of the image, and this is referred to as “determined color The color table memory 5 is controlled to be stored as a "table" and the process proceeds to step S7. As mentioned above,
In the gradation reducing apparatus, the mean square error is calculated based on the human visual characteristics. For this reason,
If the mean squared error is less than or equal to the threshold,
The reference color table indicates that it is a visually optimum color table in the gradation reduction processing of the image.

Therefore, in step S7, the gradation reduction circuit 6 supplies the multi-gradation image supplied through the inverse gamma correction circuit 2 based on the fixed color table stored in the color table memory 5. Performs gradation reduction processing of data,
This gradation reduced image data is supplied to the gamma correction circuit 10. The gamma correction circuit 10 performs gamma correction processing on the grayscale reduced image data according to the gamma characteristic of the display device that is to display the image, and outputs this to the outside through the output terminal 11. All the routines in the flowchart are finished.

The routine of steps S1 to S8 shown in the flowchart of FIG. 2 is a gradation reduction routine for one image, and when there are a plurality of images, the steps S1 to S8 are executed each time. Each routine of S8 will be repeatedly executed.

The image data supplied to the above display device is the data which is first subjected to the gamma correction processing and then is subjected to the gradation reduction processing, and also during the gradation reduction processing, The gradation reduction processing is performed using a reference color table (determined color table) detected based on human visual characteristics. Therefore, the image displayed on the display device can be visually excellent. Further, since it is possible to improve the image quality of a display image in this way, even in an image communication system or the like having a plurality of display devices having different gradation display capabilities, it is possible to display with a gradation suitable for each display device. It is possible to provide images, and it is possible to provide higher quality image communication services.

Next, a gradation reducing apparatus according to the second embodiment of the present invention will be described. The gradation reduction device according to the first embodiment described above uses visual characteristics when deriving a fixed color table for gradation reduction processing of the image, but in the second embodiment, The gradation reducing device derives a fixed color table based on the mean square error without using visual characteristics when deriving the fixed color table, and uses the fixed color table to reduce the gradation reduced image data. The visual characteristics are used when selecting the optimum one from the inside.

In the description of the tone reducing apparatus according to the second embodiment, the same operation as that of the tone reducing apparatus according to the first embodiment described above will be omitted in the drawings. The reference numerals are given and detailed description thereof is omitted.

That is, as a characteristic configuration of the gradation reducing apparatus according to the second embodiment, the mean square error (gradation reducing error) of the image is calculated by the least square method as shown in FIG. A first error detection circuit 21 for detecting, and a second error detection circuit 22 for detecting the mean square error of the image based on the visual characteristic data stored in the visual characteristic ROM 7.
And derive an optimum color table (determined color table) based on the mean square error from the first error detection circuit 21, and based on the mean square error from the second error detection circuit 22, The control circuit 23 detects the optimum image from the images subjected to the tone reduction processing, and controls the output of this as grayscale reduction image data.

The overall operation of the gradation reducing apparatus according to the second embodiment is as shown in the flowchart of FIG. 4, and the main power source of the gradation reducing apparatus is turned on. It will start and step S11
Proceed to.

In step S11, the control circuit 23 derives a fixed color table, which is an optimum color table, based on the mean square error from the first error detection circuit 21, and proceeds to step S12.

Specifically, the derivation of the fixed color table is performed without performing the inverse gamma correction process on the multi-tone image data supplied through the input terminal 1, that is,
The gamma correction process is performed on the basis of the multi-tone image data, and the control circuit 23
When the multi-tone image data is supplied through the input terminal 1, it is stored and controlled in the memory 3 as it is.

When the multi-gradation image data is stored in the memory 3, the control circuit 23 reads it and supplies it to the gradation reduction circuit 6 together with the reference color data from the reference color table 4. The gradation reduction circuit 6 performs gradation reduction processing of the multi-gradation image data based on the reference color data, and executes the first gradation reduction processing.
Is supplied to the error detection circuit 21. First error detection circuit 2
1 detects the mean square error of the image subjected to the gradation reduction processing, and supplies this to the control circuit 23. Control circuit 2
3 compares the mean square error with a predetermined threshold when the mean square error is detected, and when the mean square error is equal to or larger than a predetermined threshold, the vector quantization codebook such as the LBG algorithm. The reference color table 4 is slightly changed using the creation algorithm, and the gradation reduction process is performed again using the changed reference color table. If the mean squared error is less than or equal to a predetermined threshold value, the reference color table indicating the mean squared error is set to the “determined color table”.
Is stored in the color table memory 5.

The gradation reduction apparatus according to the second embodiment is based on the "determined color table" detected in this way, and the steps S12 to S1 described below are performed.
The routine of No. 8 is executed to detect the optimum gradation-reduced image data.

That is, in step S12, the control circuit 23 determines whether or not the image data is supplied to the gradation reducing device by detecting the presence or absence of the data supply from the inverse gamma correction circuit 2. If No, the step S12 is repeated until the image data is supplied, and if Yes, the process proceeds to step S13.

In step S13, the inverse gamma correction circuit 2
However, the gamma correction process has already been performed, and the inverse gamma correction process is performed on the multi-gradation image data supplied through the input terminal 1 to perform the gamma correction process on the multi-gradation image data. The process returns to the non-existing state and proceeds to step S14.
The multi-tone image data that has been subjected to the inverse gamma correction process is
The image is temporarily stored in the memory 3 as an original image to be subjected to the gradation reduction processing.

Next, in step S14, the gradation reduction circuit 6 uses the fixed color table stored in the color table memory 5 to reduce the gradation of the multi-gradation image data stored in the memory 3. The gradation-reduced image data that has been subjected to the gradation reduction processing is stored in the memory 3 as the original image data for selecting the optimum gradation-reduced image data, which will be described later. Error detection circuit 22
And the process proceeds to step S15.

In step S15, the second error detection circuit 22 detects the mean square error using the visual characteristic weighting coefficient stored in the visual characteristic ROM 7,
This is supplied to the control circuit 23 and the process proceeds to step S16.
Here, in the gradation reduction device, as will be described below, each pixel value of the gradation reduced image data is set as the fixed color table with the gradation reduced image data that has been subjected to the gradation reduction processing first as a reference. When the average squared error based on the above-mentioned visual characteristics is detected, the averaged squared error is compared with a predetermined threshold value each time, and is less than or equal to a predetermined threshold value. The gradation-reduced image data indicating the mean square error of is output as the gradation-reduced image data of the image.

Therefore, in step S16, the control circuit 23 determines whether or not the mean square error detected in step S15 is less than or equal to a predetermined threshold value. If No, the process proceeds to step S18, Yes. In the case of, it progresses to step S17. Next, in step S18, the control circuit 23 sets each pixel value of the gradation-reduced image data stored in the memory 3 within the range of colors on the fixed color table stored in the color table memory 5. To make a slight change,
This is supplied to the second error detection circuit 22 and step S1
Return to 5.

In step S15, as described above, the second error detection circuit 22 uses the weighting coefficient according to the visual characteristic stored in the visual characteristic ROM 7 to finely change the gradation of each pixel value. The mean square error of the reduced image data is detected, supplied to the control circuit 23, and the process proceeds to step S16. The control circuit 23 compares the mean square error with a predetermined threshold value.

As described above, in the gradation reduction device, by repeating the processing of step S15, step S16 and step S18, each pixel value of the gradation reduced image data is minutely changed to obtain the respective mean square errors. To detect. Then, when the mean square error becomes equal to or less than the predetermined threshold value, the control circuit 23 determines “Yes” in step S16 and proceeds to step S17.

The fact that the mean square error detected on the basis of the visual characteristics is equal to or less than the predetermined threshold indicates that the gradation reduced image data is good for human vision. . Therefore, in step S17, the control circuit 23 outputs the gradation-reduced image data of the mean square error that is equal to or less than the predetermined threshold value as the gradation-reduced image data of the image. The gradation-reduced image data is supplied to the gamma correction circuit 10, and the gamma correction circuit 10 performs gamma correction processing corresponding to the gamma characteristic of the display device that is to display the image,
This is output to the outside through the output terminal 11, and the whole routine of the flowchart is finished.

The routine of steps S11 to S18 shown in the flow chart of FIG. 4 is a routine for reducing the gradation of one image, and when there are a plurality of images, the above steps are executed each time. Each routine of S11 to step S18 will be repeatedly executed.

As described above, the gradation reduction device makes each pixel value of the gradation-reduced image data, which has been gradation-reduced using the fixed color table, minute within the range of colors on the fixed color table. Along with the change, a mean squared error based on human visual characteristics is detected for each change, and gradation reduced image data in which the mean squared error is equal to or less than a predetermined threshold is output as gradation reduced image data of the image. Because I am trying to
It is possible to supply gradation-reduced image data that reflects human visual characteristics to a display device. Therefore, it is possible to obtain substantially the same effect as that of the gradation reduction device according to the first embodiment described above, such as the improvement of the image quality of the display image.

Next, a gradation reducing apparatus according to the third embodiment of the present invention will be described. The gradation reduction devices according to the above-described first and second embodiments use visual characteristics when deriving a fixed color table for gradation reduction processing of the image,
Alternatively, the visual characteristic is used when the optimum one is selected from the gradation-reduced image data whose gradation is reduced by the fixed color table. However, the gradation reduction according to the third embodiment is performed. The device uses visual characteristics when deriving the fixed color table and when selecting the optimum one from the gradation-reduced image data.

In the description of the gradation reducing device according to the third embodiment, the parts similar to those of the gradation reducing devices according to the above-described first and second embodiments are shown in the drawings. The same reference numerals are given above, and detailed description thereof is omitted.

That is, as a characteristic configuration of the gradation reducing apparatus according to the third embodiment, as shown in FIG. 5, the average of the image is calculated based on the visual characteristic data stored in the visual characteristic ROM 7. A first error detection circuit 21 for detecting a squared error (gradation reduction error), and a second error detection circuit 22 for detecting a mean squared error of the image based on the visual characteristic data stored in the visual characteristic ROM 7. The optimum definite color table is derived based on the mean square error from the first error detection circuit 21, and gradation reduction processing is performed based on the mean square error from the second error detection circuit 22. It has a control circuit 25 for detecting an optimum image from the images and controlling the output as gradation reduction image data.

The overall operation of the gradation reducing apparatus according to the third embodiment is as shown in the flowchart of FIG. 6, and the main power source of the gradation reducing apparatus is turned on. This will start and step S21
Proceed to.

At step S21, the control circuit 25
Detects whether there is multi-gradation image data supplied to the inverse gamma correction circuit 2 to determine whether or not the image data has been supplied to the gradation reduction device. If No, the image data is The step S21 is repeated until it is supplied, and if Yes, the process proceeds to step S22.

In step S22, the inverse gamma correction circuit 2
However, the multi-gradation image data is subjected to the gamma correction processing by performing the inverse gamma correction processing on the multi-gradation image data which has been already subjected to the gamma correction processing, which is supplied through the input terminal 1. It is returned to the non-existing state, this is once stored in the memory 3, and the process proceeds to step S23. Next, step S23
Then, the gradation reduction circuit 6 performs the gradation reduction processing of the multi-gradation image data using the reference color table 4, and proceeds to step S24.

In step S24, the first error detection circuit 8 detects the mean square error of the image subjected to the gradation reduction processing based on the reference color table data based on the visual characteristic data stored in the visual characteristic ROM 7. , And supplies this to the control circuit 25, and proceeds to step S25.

Here, in the gradation reducing apparatus, as described below, the reference color table 4 is changed, and each time, the above-mentioned mean square error is detected and compared with a predetermined threshold value. A reference color table showing a mean square error equal to or less than a threshold is used as a “determined color table” for gradation reduction processing of the image. For this reason,
In step S25, the control circuit 25 determines whether or not the mean square error detected in step S24 is equal to or smaller than a predetermined threshold value. If No, the process proceeds to step S31, and if Yes, the process proceeds to step S26. move on.

In step S31, the control circuit 25 slightly changes the reference color table 4 by using a vector quantization codebook creation algorithm such as the LBG algorithm described above, and returns to step S23. When it is determined No in step S25, the reference color table is changed in step S31, and the process returns to step S23, the gradation reduction circuit 6 in turn determines the memory 3 based on the changed reference color table. Performs gradation reduction processing of the multi-gradation image data stored in step S24.
Proceed to. Then, in step S24, the first error detection circuit 8 detects the mean squared error weighted by the visual characteristic, and thereby the gradation-reduced image of the image is reduced based on the changed reference color table. An error is detected, the mean square error is supplied to the control circuit 25, and the process proceeds to step S25. In step S25, the control circuit 25 determines whether or not the mean square error detected in step S24 is equal to or less than a predetermined threshold value,
If No, the process proceeds to step S31, and if Yes, the process proceeds to step S26.

As described above, the gradation reducing apparatus repeats the processes of steps S23 to S25 and step S31 a predetermined number of times, thereby slightly changing the reference color table and the mean square error at that time is equal to or less than the predetermined threshold value. Is determined. Then, when the mean square error becomes equal to or less than the predetermined threshold value, the process proceeds to step S26.

Next, the fact that the mean square error detected on the basis of human visual characteristics is below a predetermined threshold means
The reference color table relating to the mean square error indicates that it is the optimum color table for performing the gradation reduction processing of the image. Therefore, the control circuit 25 causes the step S2
In 6, the reference color table related to the mean square error that is equal to or less than the predetermined threshold value is stored in the color table memory 5 as a “fixed color table” used when performing gradation reduction processing on the image, and the process proceeds to step S27. . This allows
In the color table memory 5, a reference color table most suitable for the image is stored as a fixed color table in view of human visual characteristics.

Next, the gradation reduction device according to the third embodiment selects optimum gradation-reduced image data as described below, based on the determined color table thus detected. It is like this.

That is, in step S27, the gradation reduction circuit 6 uses the determined color table of the image stored in the color table memory 5 to store the memory 3 in the memory 3.
The gradation reduction processing is performed on the multi-gradation image data stored in, the gradation reduction image data is supplied to the control circuit 25, and the process proceeds to step S28. When the gradation reduction image data is supplied, the control circuit 25 stores and controls the gradation reduction image data in the memory 3 and supplies it to the second error detection circuit 22.

In step S28, the second error detection circuit 22 uses the weighting coefficient according to the visual characteristic stored in the visual characteristic ROM 7 to perform the gradation reduction processing by the fixed color table and reduce the formed gradation. The mean square error of the image data is detected, supplied to the control circuit 25, and the process proceeds to step S29. Here, in the gradation reduction device, as described below, each pixel value of the gradation reduction image data is set with reference to the gradation reduction image data that is first subjected to the gradation reduction processing by the fixed color table. , Within the range of colors in the above-mentioned fixed color table, it is slightly changed, each time the above-mentioned mean squared error is detected, and the gradation reduction image data showing the mean squared error below a predetermined threshold value It is designed to be output as tone-reduced image data.

Therefore, in step S29, the control circuit 25 determines whether or not the mean square error is less than or equal to a predetermined threshold value. If No, the process proceeds to step S32, and if Yes, the process proceeds to step S30. move on. Next, in step S32, the control circuit 25 reads the gradation reduced image data stored in the memory 3, and the respective pixel values of the gradation reduced image data are stored in the color table memory 5. A slight change is made within the range of colors on the fixed color table, this is supplied to the second error detection circuit 22, and the process returns to step S28.

In step S28, as described above, the second error detection circuit 22 uses the weighting coefficient according to the visual characteristic stored in the visual characteristic ROM 7 to gradationally change each pixel value slightly. The mean square error of the reduced image data is detected, supplied to the control circuit 25, and the process proceeds to step S29. Then, in step S29, the control circuit 25 compares the mean square error with a predetermined threshold,
If it is less than the threshold value, the process proceeds to step S30. If it is more than the threshold value, each pixel value of the gradation reduced image data is further changed within the color range on the fixed color table in step S32, and the process returns to step S28.

As described above, the gradation reducing apparatus detects the mean square error by slightly changing each pixel value of the gradation reduced image data by repeating the processing of step S28, step S29 and step S32. To do. Then, when the mean square error of the gradation reduced image data becomes equal to or less than a predetermined threshold value, the process proceeds to step S30.

As described above, the mean square error is detected based on the human visual characteristics. Therefore, the fact that the mean square error is equal to or less than the predetermined threshold value means that the gradation-reduced image data indicating the mean square error is good for human vision. Therefore, in step S30, the control circuit 25 selects and outputs the gradation-reduced image data indicating the mean square error equal to or less than the predetermined threshold value as the gradation-reduced image data of the image.
The gradation-reduced image data is subjected to gamma correction processing by the gamma correction circuit 10 according to the gamma characteristic of the display device that is to display the image, and the output terminal 11
Is output to the outside via, and all the routines of the flowchart are finished.

Note that the routine of steps S21 to S32 shown in the flowchart of FIG. 6 is a routine for reducing the gradation of one image, and when there are a plurality of images, the above steps are performed each time. Each routine of S21 to step S32 will be repeatedly executed.

As described above, the gradation reducing device uses the visual characteristics when deriving the color table and when selecting the gradation-reduced image data whose gradation is reduced using this color table. Therefore, it is possible to form gradation-reduced image data that reflects human visual characteristics more than the gradation-reduction devices according to the first and second embodiments. Therefore, the effects described in the description of the gradation reducing devices according to the first and second embodiments can be made more prominent.

In the description of each of the above embodiments, the gradation reducing apparatus and the gradation reducing method according to the present invention are applied to multi-tone image data in which RGB is 24 bits, and RGB is 3 bits and 3 respectively. It was decided to apply this to a gradation reduction device that performs gradation reduction processing on 8-bit image data of 2 bits and 2 bits in total, but this is just an example.
As long as B is applied to a gradation reduction device that performs gradation reduction processing on a total of 16 bits of image data of 6 bits, 6 bits, and 4 bits, as long as it does not deviate from the technical idea of the present invention. Of course, various changes are possible.

[0086]

The gradation reduction apparatus and gradation reduction method according to the present invention can form gradation-reduced image information that reflects human visual characteristics. Therefore, it is possible to improve the image quality of the display image related to the gradation reduction image information.

Further, even in an image communication system in which display devices having various gradation display capabilities are mixed, it is possible to provide an image communication service of higher image quality.

[Brief description of drawings]

FIG. 1 is a block diagram of a gradation reducing apparatus according to a first embodiment to which a gradation reducing apparatus and a gradation reducing method according to the present invention are applied.

FIG. 2 is a flowchart showing an operation of the gradation reduction device according to the first embodiment.

FIG. 3 is a block diagram of a gradation reducing apparatus according to a second embodiment to which a gradation reducing apparatus and a gradation reducing method according to the present invention are applied.

FIG. 4 is a flowchart showing an operation of the gradation reducing device according to the second embodiment.

FIG. 5 is a block diagram of a gradation reducing apparatus according to a third embodiment to which a gradation reducing apparatus and a gradation reducing method according to the present invention are applied.

FIG. 6 is a flowchart showing an operation of the gradation reducing device according to the third embodiment.

FIG. 7 is a diagram for explaining a conventional gradation reduction process.

[Explanation of symbols]

 1 Input terminal for multi-tone image data 2 Reverse gamma correction circuit 3 Memory 4 Reference color table 5 Color table memory 6 Gradation reduction circuit 7 Visual characteristic ROM 8 First error detection circuit 9, 23, 25 Control circuit 10 Gamma correction Circuit 11 Grayscale reduced image data output terminal

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Taichi Nakamura 3-3-3 Toyosu, Koto-ku, Tokyo NTT Data Communications Corporation

Claims (12)

[Claims] 1. A gradation reduction means for performing gradation reduction processing on multi-gradation image information supplied from the outside to form gradation reduced image information, and the gradation reduction means based on human visual characteristics. Error detection means for detecting the gradation reduction error of the gradation reduction image information formed by the above, and the color table selected according to the gradation reduction error from the error detection means to obtain the multi-gradation image information of the image. A gradation reduction device comprising: a control unit that controls the gradation reduction unit to perform a gradation reduction process. 2. A reference color table made up of color information used as a reference when performing gradation reduction processing of the image, wherein the control means changes the reference color table, and for each reference color table. The gradation reduction means is controlled to perform the gradation reduction processing, and the gradation reduction error for each reference color table is detected by the error detection output from the error detection means, and the gradation reduction error is minimized. 2. The floor according to claim 1, wherein the gradation reduction means is controlled so as to perform gradation reduction processing on the multi-gradation image information of the image by using a reference color table indicating the value of as the color table. Adjustment device. 3. A multi-grayscale image in a non-gamma-corrected state is obtained by performing an inverse gamma-correction process on arbitrary multi-grayscale image information that has been gamma-corrected and is externally supplied. 3. The gradation reduction device according to claim 1, further comprising an inverse gamma correction unit that corrects information and supplies it to the gradation reduction unit. 4. A gradation table is subjected to gradation reduction processing based on the color table used for gradation reduction processing and arbitrary multi-gradation image information supplied from the outside to form gradation reduction image information. And the gradation reduction image information is changed within the range of the color information of the color table, and the gradation reduction error of each gradation reduction image information is detected based on human visual characteristics, Gradation reduction image information selecting means for selecting and outputting gradation reduction image information indicating the gradation reduction error of 1. 5. A gradation reduction of a reference color table composed of color information used as a reference in gradation reduction processing and arbitrary multi-gradation image information supplied from the outside while changing the reference color table. Color table variable gradation reduction means for processing and error detection means for detecting gradation reduction error of gradation reduced image information for each of the reference color tables that has been subjected to gradation reduction by the color table variable gradation reduction means. And the gradation reducing means detects a reference color table showing a minimum value of the gradation reduction error detected by the error detecting means from the changed respective reference color tables, The gradation reduction device according to claim 4, wherein a gradation reduction process is performed on the multi-gradation image information by using a color table as the color table. 6. The gradation-reduced image information selection means stores each pixel value of the gradation-reduced image information formed by the gradation-reduction processing based on the color table by the gradation-reduction means in the color table. Pixel value changing means for changing within the range of color information, and an error for detecting a gradation reduction error of gradation reduction image information formed by gradation reduction processing based on the color table based on human visual characteristics A detection unit, a gradation reduction error of gradation reduction image information formed by gradation reduction processing based on the color table, and a floor of each gradation reduction image information in which each pixel value is changed by the pixel value changing unit. The error detecting means is controlled so as to detect each gradation reduction error, and the gradation reduction image information indicating the minimum gradation reduction error is selected from among the gradation reduction errors detected by the error detecting means. Output The gradation reduction apparatus according to claim 4 or 5, wherein the gradation reduction image information output means is included. 7. The gradation reduction of the gradation-reduced image information formed by the gradation reduction means by applying gradation reduction processing to multi-gradation image information supplied from the outside to form gradation-reduced image information. An error is detected based on human visual characteristics, a color table is selected according to the detected gradation reduction error, and a gradation reduction process is performed on the multi-gradation image information of the image by this color table. Reduction method. 8. When the color table is selected, the reference color table including color information serving as a reference when performing gradation reduction processing of the image is changed, and the multi-gradation is performed for each of the reference color tables. Gradation reduction processing is performed on image information, a gradation reduction error for each of the reference color tables is detected, and the reference color table showing the minimum value of the gradation reduction error is selected as the color table. Claim 7
The described gradation reduction method. 9. The gradation reduction image information is formed by performing gradation reduction processing of multi-gradation image information supplied from the outside according to a color table used in the gradation reduction processing. While changing within the range of the color information of the color table, the gradation reduction error of each gradation reduction image information is detected based on the human visual characteristics, and the gradation reduction image information showing the minimum gradation reduction error is selected. A gradation reduction method that outputs 10. The color table used in the gradation reduction processing is a gradation reduction processing of multi-gradation image information supplied from the outside while changing a reference color table composed of reference color information, A gradation reduction error of gradation reduction image information for each reference color table is detected, and a reference color table showing the minimum gradation reduction error is selected from the changed reference color tables. 10. The gradation reduction method according to claim 9, wherein: 11. When selecting the gradation reduction image information, each pixel value of the gradation reduction image information formed by gradation reduction processing based on the color table is set to a range of the color information of the color table. The gradation reduction error of each gradation reduction image information is detected based on human visual characteristics while showing the smallest gradation reduction error among the detected gradation reduction errors. The gradation reduction method according to claim 9 or 10, wherein gradation reduction image information is selected. 12. Multi-gradation image information supplied from the outside is subjected to gradation reduction processing to form gradation reduction image information, and a gradation reduction error of the gradation reduction image information is based on human visual characteristics. Is detected, the gradation table is selected according to the gradation reduction error, and the gradation reduction processing is performed on the multi-gradation image information of the image. , The gradation reduction error of each gradation reduction image information is detected based on the human visual characteristics while changing within the range of the color information of the color table, and the gradation reduction image information showing the minimum gradation reduction error is displayed. A gradation reduction method to select and output.