JPH06343131A - Multidimensional color image compression system - Google Patents

Abstract

PURPOSE:To obtain a high-efficiency compression system for a color image by obtaining feature points of the equal-hue line of a hue function and feature points of the equal-luminance line of the luminance function in the area encircled with the equal-hue line and using the positions, luminance values, and chromaticity values of feature points of image information for transmission, recording, and image restoration. CONSTITUTION:This system consists of an image signal source 1, an A/D converter 2, a write control part 3, a hue contour tracer 5a, a luminance contour tracer 5b, an encoding and sending circuit 6, a receiving and decoding circuit 8, a luminance function reproducing device 9, etc. Then the whole image to be compressed and expanded is sectioned with the equal-hue line, and the positive and negative maximum points of the curvature of the equal-hue line and the positions and attributes (luminance and chromaticity) of important feature points which are not included in linear approximation are transmitted so that the equal-hue line can be reproduced on an expansion side. The luminance contour in its internal area is found and the positive and negative maximum points of the curvature of the equal-hue line and the positions and attributes (luminance and chromaticity) of important feature points which are not included in linear approximation are transmitted so that the luminance contour can be reproduced on the expansion side.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention guarantees a multi-dimensional color image compression system, in particular, pixel correspondence on the compression side and decompression side such as image transmission between different models, or frame correspondence on the time axis. The present invention relates to a multi-dimensional color image compression method that can obtain compressed image data that can perform highly efficient image expansion even in a system that does not operate.

[0002]

2. Description of the Related Art Various methods have been conventionally proposed for compression of image information. For example, when linear quantization (equal quantization) means is adopted in which the signal level is evenly divided for each sampled value of the digitalized image signal and the values included in each range are replaced with one representative value. It is said that 6 bits (64 gradations) to 8 bits (256 gradations) are generally required for a natural image when the difference between the representative point and the original value is not known.
In order to record a signal obtained by digitizing the image signal by the uniform quantization as described above, it is necessary to handle a large amount of information as described above for each sample value.

Therefore, even if the signal is encoded with a smaller amount of information, it is sensitive to the change in the part where the signal changes little and there is some error in the part where the signal changes drastically. It is difficult to detect the human visual and auditory characteristics, or by utilizing the correlation on the spatiotemporal axis of the information signal that is the target of recording, for example, after dividing the image into pixels, each pixel Use the height of the adjacent correlation of the luminance values to transmit a small number of the approximate values of the original information, or to transmit the pixel-to-pixel difference or the frame-to-frame difference, or to use the fact that there are few high frequency components. The digital data is compressed by applying various high-efficiency coding schemes that reduce the amount of information for each sample by reducing the frequency elements. Record, It has been conventionally practiced to restore an image by transmitting and transmitting, and by reproducing and receiving the digital data in which the amount of data is compressed as described above, and then expanding the data. As is well known.

[0004]

In the above-mentioned conventional general compression method of image information, it is important to restore the decomposed pixels satisfactorily. In many cases, the condition is that the number of pixels in the images (decompressed images) is the same. Therefore, when compression / decompression operation is performed between images with different numbers of pixels, pixel interpolation after decompression is performed separately. However, in the conventional compression method of image information, only true effective information is extracted and not restored, and it is a physical image constituent element to some extent. It means that the method depends on.

By the way, as described above, two pixels having different numbers of pixels are used.
As an example of a case where the pixel densities of two images are extremely different, for example, when an image captured by the image capturing device is used as a printing plate, the pixel density of the image captured by the image capturing device is one screen. The maximum pixel density per image is 500 (500 x 500), whereas the pixel density of an image on an electronic plate making machine is (thousands x thousands) per screen.
As described above, since it is an order of magnitude larger than the image obtained by the image pickup by the image pickup apparatus described above, aliasing occurs due to pixel enlargement even if the compression / expansion method of image information corresponding to the pixel as described above is not performed at all. Do
Further, when the interpolation is performed without performing the pixel enlargement as described above, a large interpolation area is set to a known data.
However, the image quality deterioration cannot be avoided due to the interpolation distortion because the weighted average value of the data is applied. Contrary to the above, in the case where the pixel density of the original image is (several thousands × thousands) per screen, the correlation between adjacent pixels is extremely high. Although it can be compressed, the original image and the restored image (decompressed image) as described above
The conventional image information compression method, which requires that the numbers of pixels are the same, has a drawback that the compression rate cannot be increased. In order to solve the above-mentioned drawbacks, there is a method of displaying the brightness function of an image in a vector manner, and a method of transmitting and restoring characteristic points of equal brightness lines has been proposed as described later. However, since the most precise restoration is required for the facial expression of a human face, if the density of isoluminance lines is determined correspondingly, the problem remains that the compression rate cannot be increased in the entire image.

[0006]

SUMMARY OF THE INVENTION According to the present invention, with respect to image information including a two-dimensionally distributed color in a still image, the positive and negative maximum points of the curvature of the equal hue line of the color correlation number or the equal hue line of the color correlation number. The error between the straight line approximated by the straight line and the equal hue line of the above-mentioned color correlation number exceeds a predetermined threshold value, and the curvature of the equal luminance line of the luminance function in the area included in the above equal hue line. The positive and negative maximum points of, or a straight line obtained by linearly approximating the isoluminance line of the luminance function, and the error between the isoluminance line of the luminance function described above, the point that exceeds a predetermined threshold value as a feature point of the image , The position of the characteristic point of the image and the luminance value, the chromaticity value (or the color difference value) are used for transmission, recording, and image restoration, and other than the characteristic point by the interpolation surface determined by a plurality of neighboring characteristic points at the time of expansion. Brightness information and color information of each pixel A multi-dimensional color image compression method capable of obtaining compressed image data capable of achieving the above, and a three-dimensional distribution in which a time axis is also added to image information including a two-dimensionally distributed color in a still image. Regarding the image information of, the difference between the positive and negative maximum points of the curvature of the color hue correlation surface of the equal hue plane, or the error between the plane obtained by approximating the color hue correlation plane of the color hue correlation plane and the color hue correlation plane of the color hue correlation coefficient is predetermined. Point exceeding the threshold value, and the positive and negative maximum points of the curvature of the equiluminance surface of the luminance function in the area included in the above-mentioned equal hue surface, or a plane that is a plane approximation of the equal luminance surface of the above-mentioned luminance function and The point where the difference between the luminance function and the isoluminance surface exceeds a predetermined threshold is defined as the characteristic point of the image, and the position of the characteristic point of the image and the luminance value, the chromaticity value (or the color difference value)
Are used for transmission, recording, and image restoration, and compressed image data that can determine the luminance information and color information of pixels other than the feature points by the interpolated solid determined by a plurality of neighboring feature points when decompressing are obtained. The multi-dimensional color image compression method as described above and the above-mentioned two-dimensionally distributed image information in which a time axis is also added to the image information including the two-dimensionally distributed colors in the still image, Targeting a plurality of sets of image information that are dimensionally distributed, the positive and negative maximum points of curvature of the equal hue line of the color correlation number of each set, or a straight line that approximates the equal hue line of the color correlation number, and the hue described above. The error between the function and the isochromatic line exceeds a predetermined threshold value, and the positive and negative maximum points of the curvature of the isoluminous line of the luminance function in the region included in the aforementioned isochromatic line, or Isoluminance line of the luminance function The error between the straight line approximated by a straight line and the isoluminance line of the brightness function described above is defined as the image feature point at a point exceeding a predetermined threshold, and the position and brightness value of the image feature point and the chromaticity value are described. (Or color difference value)
It is used for image restoration and it is possible to obtain compressed image data that can determine the luminance information and color information of pixels other than the feature points by the interpolated solid that is determined by a plurality of neighboring feature points when decompressing. Regarding the color image compression method and the image information including the two-dimensionally distributed color in the still image, the positive and negative maximum points of the curvature of the equiluminance line of the luminance function in the region included in the equal hue plane, or the luminance A straight line obtained by linearly approximating the isoluminance line of the function and the error between the isoluminance line of the above-mentioned luminance function and the position of the above-mentioned feature point of the image as a feature point of the image at a point exceeding a predetermined threshold value. The luminance value and the chromaticity value (or color difference value) are used for transmission, recording, and image restoration, and the luminance information of pixels other than the feature point is determined by the interpolation plane determined by a plurality of neighboring feature points during expansion.
A multi-dimensional color image compression method capable of obtaining compressed image data capable of determining color information, and three-dimensional image information including a two-dimensionally distributed color in a still image and a time axis The image information that is distributed in a distributed manner, the positive and negative maximum points of the curvature of the equiluminance surface of the luminance function in the area included in the equihue solid, or a plane that is a plane approximation of the equiluminance surface of the luminance function described above. The point where the error of the luminance function from the isoluminance surface exceeds a predetermined threshold is set as the characteristic point of the image, and the position of the characteristic point of the image and the luminance value and chromaticity value (or color difference value) are It is used for transmission, recording, and image restoration so that compressed image data can be obtained that can determine the luminance information and color information of pixels other than the feature points by the interpolated solid that is determined by a plurality of neighboring feature points when decompressing. Many The original color image compression method, and the two-dimensional distribution from the three-dimensionally distributed image information in which a time axis is added to the image information including the two-dimensionally distributed colors in the still image. For multiple sets of image information, the positive and negative maximum points of curvature of the equiluminance line of the luminance function in the region included in the equal hue plane of each of the above sets, or the equiluminance line of the above luminance function is linearly approximated. The difference between the straight line and the isoluminance line of the brightness function described above is defined as a feature point of the image at a point exceeding a predetermined threshold value, and the position of the feature point of the image and the brightness value, the chromaticity value, ( (Or color difference value) is used for transmission, recording, and image restoration, and compression such that luminance information and color information of pixels other than the feature points can be determined by an interpolation solid that is determined by a plurality of neighboring feature points when decompressing. So that the image data can be obtained In the multi-dimensional color image compression method, in particular, the multi-dimensional color image compression method that improves the quality of the restored image by increasing the density of the equi-luminance lines or the equi-luminance surface in the hue region of human skin color is provided. .

[0007]

Operation: Regardless of whether the pixel density of an image to be image-processed is high or low, only the characteristic points of the image are extracted to obtain image data in which the image information is compressed. Instead of performing pixel restoration from the above-mentioned image data, so that a new image can be drawn on another pixel density plane,
Regarding the image information including the two-dimensionally distributed colors and the above-mentioned three-dimensionally distributed image information including the time axis as well, the equal hue lines of the color correlation numbers and the positive and negative maximum points of the curvature of the equal hue surface, Or, the error between the equal hue line or the equal hue plane of the color correlation number approximated by a straight line or a plane and the equal hue line or the equal hue plane of the above-mentioned color correlation number exceeds a predetermined threshold value, and An equal-intensity line of the luminance function in the region included in the equal-hue line or the equal-hue surface, a positive or negative maximum point of the curvature of the equal-luminance surface, or an equal-luminance line or the equal-luminance surface of the above-mentioned luminance function is a straight line or a plane. The difference between the approximated value and the isoluminance line or isoluminance surface of the above-mentioned luminance function is a point that exceeds a predetermined threshold value as an image characteristic point, and the position and luminance value of the characteristic point of the image and the color The degree value (or color difference value) is used for transmission, recording, and image restoration, and the Luminance information of the pixels other than the characteristic point by interpolation plane or interpolative stereoscopic determined by the feature points,
Try to determine the color information. In addition, among the three-dimensionally distributed image information including the time axis in the image information including the two-dimensionally distributed colors, a plurality of sets of the above-mentioned two-dimensionally distributed image information are targeted, and The positive and negative maximum values of the curvature of the equal hue line of the color correlation number of each set, or the error between the straight line approximating the equal hue line of the color correlation number of each set and the equal hue line of the color correlation number is predetermined. Points that exceed the threshold
And the positive and negative maximum points of the curvature of the equiluminance line of the luminance function in the area included in the above-mentioned equal hue line, or a straight line obtained by linearly approximating the equal-luminance line of the above-mentioned luminance function and the equal-luminance line of the above-mentioned luminance function A point where the error between and exceeds a predetermined threshold is used as an image feature point, and the position of the image feature point and the luminance value and chromaticity value (or color difference value) are used for transmission, recording, and image restoration. During the expansion, the luminance information and the color information of the pixels other than the feature points are determined by the interpolated solids determined by the plurality of feature points in the vicinity. Further, when determining the luminance information and color information of pixels other than the feature points as described above, when the equi-hue region is the hue of human skin color, the density of the equi-luminance line and the equi-luminance surface is increased. To do so.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Specific contents relating to the multi-dimensional color image compression system of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 is a block diagram of a multi-dimensional color image compression / expansion method configured to include the multi-dimensional color image compression method of the present invention, and FIGS. 2 and 3 are used for the multi-dimensional color image compression method of the present invention. It is a block diagram which shows the structural example of the contour line tracer. Further, FIG. 4 is a block diagram showing a configuration example of the luminance function reproducing device, FIG. 5 is a diagram for explaining a method of detecting a luminance contour line and a method of determining characteristic points, and FIG. 6 is a configuration of the luminance function reproducing device. FIG. 7 is a diagram used for explaining the principle and operation principle, FIG. 7 is a diagram used for explaining the luminance function reproduction, FIG. 8 is a diagram used for explaining a surface interpolation operation for the luminance function reproduction, and FIG. 9 is three-dimensionally It is a figure used for description in the case of determining the brightness information of the pixels other than the feature points by the interpolated solid determined by the feature points detected from the distributed image information.

According to the present invention, image information of a color image having luminance [brightness] and chromaticity [hue and saturation (saturation)] is two-dimensionally distributed to be compressed and expanded. The present invention relates to a multi-dimensional color image compression method in which information and image information that is to be compressed and expanded and that is two-dimensionally distributed include three-dimensionally distributed image information including a time axis. Taking image information that is two-dimensionally distributed as an example, the image information has two-dimensional distributions of luminance, hue, and saturation information, and a two-dimensional luminance function, a two-dimensional color correlation number, and a two-dimensional color correlation number, respectively. Form a dimensional saturation function. By the way, when a human distinguishes a display object from the general image information that is two-dimensionally distributed as described above, it is often divided into a figure and a ground. The boundaries are often hue boundaries. This is because a single object often has the same color. However,
Since the shape of the display object to be identified is indefinite, brightness and saturation are randomly generated depending on the illumination state. Therefore, it cannot be said that the brightness and the saturation are constant within the object.

Therefore, in the multi-dimensional color image compression method of the present invention, the entire image to be compressed and decompressed is first divided by equal hue lines so that the equal hue lines can be reproduced on the decompression side. , The position and attributes (luminance, chromaticity) of the important feature points of the positive and negative maximum points of the curvature of the above-mentioned hue line and points deviating from the linear approximation are transmitted, and the luminance contour line is obtained for the internal region, The position and attributes (luminance and chromaticity) of important feature points of the positive and negative maximum points of the curvature of the above-mentioned hue line and points deviating from the linear approximation should be transmitted so that the luminance contour line can also be reproduced on the extension side. I have to.

As described above, there are two types of target functions for extracting feature points in the multi-dimensional color image compression method of the present invention, the color correlation number and the luminance function. And the intensity function are similar,
In the following description, the brightness function will be taken up as a representative example. Now, let z be the brightness in the monochrome still image,
Further, assuming that the horizontal direction of the screen is x and the vertical direction is y, the image can be generally expressed by the following equation (1). z = f (x, y) (1) Further, if the time axis is t, the moving image can be expressed by the following equation (2). z = f
(X, y, t) (2) Here, if f is a multi-dimensional function, the brightness z in the image is expressed by the following equation (3).

[0012]

[Equation 1]

It can be said that transmitting an image reproduces the above-described luminance function determined on the image transmitting side on the image receiving side. However, in general digital image transmission, the luminance function is analyzed analytically. All of the table values are transmitted without using it as a so-called table function. On the other hand, in conventional general compression transmission, high-efficiency coding using the adjacent correlation of the table value itself is performed, or the table after orthogonal transformation is used.
Although the same measures are taken for the bull value, there are few conventional compression transmission systems that directly extract the feature value of the function from the analytical processing related to the luminance function.

It can be said that the multi-dimensional color image compression system of the present invention belongs to an image compression / decompression system in which the feature value of the function is directly extracted from the analytical processing relating to the luminance and the color correlation number. However, according to the multi-dimensional color image compression method of the present invention, two-dimensional brightness and hue information in a still image, or two-dimensional brightness and three-dimensional brightness and hue information including a time axis in the hue information, etc. For the image information of (3), the iso-hue line H and the equi-luminance line (contour line) L therein (illustrated in FIG. 5C) (in the case of three-dimensional luminance information, iso-luminance, hue plane) are extracted. To do. The inner region of the equal hue line H is the hue of a human skin color (in the color space, the hue of the red primary color in the three primary colors of additive mixture is 0 degree, and the hue of the blue primary color is 1 degree).
When the hue of the primary color of 20 degrees and green is represented as 240 degrees, when the hue is in the range of 10 degrees to 30 degrees), the isoluminance linear density is extracted so as to be high. The extracted positive and negative maximum points of curvature of the equal luminance, hue line (contour line) (in the case of three-dimensional luminance information, equal luminance, hue plane),
Or, the error from the straight line approximation (equal luminance, hue plane in the case of three-dimensional luminance information) of the above-mentioned equal luminance, hue line (contour line) (plane approximation in the case of three-dimensional luminance, hue information) ,
A point that exceeds a predetermined threshold value is used as a characteristic point of the image described above, and the position and luminance value and chromaticity value of the characteristic point of the image described above are used for transmission (recording) and image restoration to reproduce the image. It is possible to realize significant compression of the image information by rejecting the information of the pixels that have little effect on the By performing re-interpolation, untransmitted information (unrecorded information) can be interpolated and reproduced.

In FIG. 1 showing a multi-dimensional color image compression system of the present invention, reference numeral 1 is a signal source of an image signal to be subject to multi-dimensional color image compression / expansion, and the image signal source 1 is, for example, an image. An image pickup device (TV camera) or a VTR that generates a signal or the like can be used. Reference numeral 2 is an analog-to-digital converter, 3 is a writing controller for performing writing control when writing the digital image data output from the analog-to-digital converter 2 into the image memories 4a and 4b, and 5a is for hue. Contour tracer (see Figure 3 for a detailed configuration example)
5b is a luminance contour tracer (a detailed configuration example is shown in FIG. 2), 6 is an encoding / sending circuit, 7 is a transmission line (or recording medium), and 8 is a receiving line. Decoding circuit (or reproduction decoding circuit), 9 is a luminance function reproducing device, 1
Reference numeral 0 is a drive circuit, and 11 is a monitor receiver.

The principle of compression of image information and the principle of restoration (expansion) in the multi-dimensional color image compression system of the present invention will be described below. First, the above-mentioned characteristic points extracted from an image when image information compression is performed on a still image, that is, the equal hue line (contour line) of the color correlation number (in the case of three-dimensional hue information, the equal hue plane). From the positive and negative maximum points of the curvature of or of the above-mentioned equal hue line (contour line) {equal hue plane in the case of three-dimensional hue information} linear approximation (plane approximation in the case of three-dimensional hue information) Regarding the characteristic points of the image when the points whose error exceeds the predetermined threshold are also extracted: 1. The above-mentioned contour lines before and after the pixel a under consideration {in the case of three-dimensional hue information If the bend of the contour plane exceeds a predetermined threshold angle, it is determined that the pixel a is a feature point.
From there, in accordance with the contour line {contour surface in the case of three-dimensional hue information}, a virtual straight line between the pixel and a pixel traced in a certain direction, the pixel between the two pixels, A pixel indicating a distance exceeding a certain threshold distance is determined to be a feature point.

Next, the characteristic points of the equiluminance line of the luminance function of the internal region divided by the equal hue line are extracted. The characteristic point extraction processing of the equiluminance line of the luminance function is described above. Similar to that for isochromatic lines. That is, the positive and negative maximum points of the curvature of the isoluminance line (contour line) of the luminance function (the isoluminance surface in the case of three-dimensional luminance information), or the isoluminance line (contour line) {three-dimensional luminance information In the case of, the characteristic points of the image in the case where the error from the linear approximation (in the case of three-dimensional luminance information, the plane approximation) of the equal luminance surface} also exceeds a predetermined threshold value, 1. If the curve of the above-mentioned contour line (contour surface in the case of three-dimensional luminance information) before and after the pixel b under consideration exceeds a predetermined threshold angle, It is determined that the pixel b is a feature point 2. The detected feature point pixel and the contour line (the contour surface in the case of three-dimensional luminance information) from the pixel are traced in a certain direction. Pixels between the two pixels described above with respect to a virtual straight line between the pixels , It is determined that a pixel indicating a distance exceeding a certain threshold distance is a feature point.In addition, the inside area of the equal hue line is the hue of the human skin color (in the color space, the red primary color of the three primary colors of additive color mixture) When the hue is expressed as 0 degree, the hue of the blue primary color is 120 degrees, and the hue of the green primary color is 240 degrees, when the hue is in the range of 10 degrees to 30 degrees, the above-mentioned equal luminance linear density (equal luminance) is obtained. By increasing the areal density, it is possible to perform compression without impairing the fidelity.

Here, a specific example of the case where the characteristic points of the contour lines of the hue and the characteristic points of the contour lines of the luminance in the two-dimensional screen in an arbitrary still image are determined according to the above-mentioned criteria will be described. It should be noted that, since the characteristic point determination method of the contour line of the hue and the luminance in the two-dimensional screen in the still image is similar to the hue and the luminance as described above, the luminance inside the equal hue line is described here. A specific example of the case of extracting the feature points of the target image for image information compression as described above will be described with reference to FIG. 5 exemplifying the distribution function. 5A illustrates one of the contour lines of the brightness distribution function in the two-dimensional screen, the extraction of the contour line of the brightness distribution function in the two-dimensional screen is as follows. It is done by procedure. FIG. 5B
With reference to, a method of searching for points showing equal brightness will be described. In FIG. 5B, (1) to (4) are four pixels, and the brightness values of the pixels (1) to (4) are H, L (H is a high brightness value, L is a low brightness value), and the point S at which the first extracted value is obtained is the starting point of contour line extraction. Since the point S is in the middle of the pixel (1) and the pixel (4), the brightness value at the coordinate position is, for example, the brightness value of the pixel (1) and the pixel
You may use the proportional distribution value of the brightness value according to the brightness value of (4).

Isoluminance lines (contour lines) obtained by connecting the points having the luminance value equal to the luminance value from the starting point S point to the point S
In order to track the image, the sides (1)-(2), (2) -in the image shape indicated by the above-mentioned four pixels (1) to (4).
When the presence or absence of a point having the same brightness value (contour value) as the brightness value of the start point S is verified in the order of (3), (3)-(4), the above-mentioned contour value is the side (1)- (2) Detected at point M1 above. After the point M1 having the above-mentioned contour value is detected, the procedure described above is repeated to detect the next point M2 having the contour value. Go forward. In the extraction of the contour lines, the side where the contour lines have already intersected is excluded from the detection target (except for the start point S described above). As described above, by detecting the sequential points having a certain luminance value and the equal luminance value, one of the contour lines of the luminance distribution function in the two-dimensional screen as illustrated in FIG. One is extracted.

One of the contour lines of the luminance distribution function in the two-dimensional screen
5A exemplifying one of the two points, the path of the isoluminance line (contour line) formed by a series of isoluminance points sequentially traced from the start point S is suddenly changed. M6 is 1. of the criteria for determining the feature points described above. It is determined as a feature point according to (points M8, M14, M16, M18, M
Regarding 20 as well, 1. According to the above). After the point M8 indicating the equal brightness value is detected and the points of equal brightness value are sequentially tracked to reach the point M12, the point M10 at which the distance from the virtual straight line M8-M12 occurs is Although it is not a sudden change point in 2. Is determined as a feature point. This also applies to the points M12 and M22.
As described above, since the luminance distribution function in the two-dimensional screen is replaced with contour lines (contour lines), the image information is rejected, and the feature information of the contour lines is further aggregated. The image compression method allows transmission (recording) in a highly compressed state.

As described above, in the multi-dimensional color image compression method of the present invention, the two-dimensional luminance and the color correlation number of the image which is the object of the image information compression are the positional information, the luminance and the color of the few feature points. The image data transmitted (recorded) in a state where the image information is highly compressed by decompressing (recovering) the image information is obtained by replacing Determining the luminance and chromaticity values of the corresponding points (which do not need to correspond to the original image and pixels) and losing the influence on neighboring pixels in proportion to the distance from the feature point The law is adopted. The image information expansion as described above is performed by linear interpolation between feature points in one dimension, but the image information expansion in a multidimensional space is performed by plane interpolation or stereoscopic interpolation. Next, the principle will be described with reference to FIG. 6 by taking a case of a luminance function as a typical example.

FIG. 6 shows the characteristic points S, shown in FIG. 5A obtained as described above with reference to FIG.
M6, M8, M10, M12, M14, M16, M18, M20, M22
Contour lines (isoluminance lines) reproduced on the image memory on the reproduction side are included by using the position information (address). In FIG. 6, the contour lines (contour lines of the M group) are the characteristic points S, M6, M8, M10, M12, M14, M16, M18, M2.
It is indicated by a dotted line that sequentially connects 0 and M22. Further, in FIG. 6, a dotted line that sequentially connects the respective feature points Ni, Nj, Nk, and Nl indicates a contour line of the feature points of the N group that is different from the contour line of the feature points of the M group. Now, when the area between the contour line of the M group and the contour line of the N group is applied with the brightness value indicated by the N group, the area between the contour value of the M group and the brightness value in the area surrounded by the contour line of the M group is Due to the clear luminance step that occurs, a so-called pseudo contour phenomenon occurs, so that a high-quality decompressed image cannot be obtained. Therefore, in order to prevent the pseudo contour phenomenon as described above from occurring, the feature points belonging to the M group and the feature points belonging to the N group are represented by za = Σ (zi / ri) / Σ (1 / ri)
(A) Applies by plane interpolation, which is an approximate operation of the formula (A).

Now, considering a specific pixel a in the expanded screen, the distance from the specific pixel a to the pixel of each feature point is ri.
And the brightness value of the pixel of each feature point is zi, and α is a constant of proportionality, the brightness value zk of one feature point k among the feature points and the brightness value za of the pixel a are The relationship is represented by the following interpolation formula (4). za = zk + αkrk (4) αkrk in the second term on the right side of the equation (4) represents the difference between the luminance value of the pixel a and the luminance value zk of the feature point k, The value of αk · rk is proportional to the distance r. The brightness value of the pixel a and the brightness value zk of the feature point k described above.
There is a positive / negative difference in the luminance value αk · rk between and, which is reflected in the proportional constant α, but the total sum of α in the entire interpolation space is zero. Then, Σαk = 0 (5)
When the general interpolation formula of the luminance value za of the pixel a is obtained by deleting the term of α from the above formula (4) in consideration of the condition of the formula (5), the above formula (A) is obtained. Be done. za = Σ (zi / ri) / Σ (1 / ri) (A)

The above equation (A) is used for the brightness values zi of all feature points.
And the distance r to the pixel a that is the interpolation target is known,
It shows that the brightness values of all the non-interpolated pixels (which may not correspond to the original image) in the screen can be obtained by interpolation. However, since the expression (A) exceeds the practical calculation amount as the number of feature points increases, the interpolated luminance value is approximately calculated by the interpolated surface determined by the three neighboring feature points surrounding the uninterpolated pixel. Since one plane is determined by three points in the space as is well known, an interpolating triangle can be obtained by grouping three neighboring feature points. Therefore, each of the three feature points “N” in FIG.
i, Nj, S "," Nk, M6, S "," Nj, Nk, S "
The luminance value of the pixel in each triangle formed by the above can be applied by the plane interpolation by the approximation calculation of the above-mentioned expression (A).

Now, in order to determine the interpolation surface when performing the above-mentioned plane interpolation, it is necessary to extract three feature points in the vicinity. Next, referring to FIG. The principle of extracting one feature point will be described. In FIG. 8, pixels Mi and Mj are characteristic points on the contour line (contourance line) of the M group whose luminance value (contourance value) is M, and the pixel N
i, Nj, Nk are N such that the brightness value (contour value) is N.
It is a feature point on the contour line (isoluminance line) of the group, and further, pixels Oi, 0j, 0k are feature points on the contour line (isoluminance line) of the O group whose luminance value (contour value) is O. When the image is decompressed, the data of the characteristic point group having the above-described contour values is stored in the reproduction side image memory. In the figure, each of the contour lines described above is indicated by a different dotted line.

For any pixel other than the feature point, to find out which feature point the arbitrary pixel is close to,
For example, it can be performed as shown in FIG. In FIG. 7, when it is assumed that feature point pixels M and N exist in the vicinity of an arbitrary pixel P, adjacent pixels are spirally inspected from the arbitrary pixel P according to the path indicated by a dotted arrow in the figure. When it goes, the pixel M of the feature point is reached before the pixel N of the feature point. As a result, the arbitrary pixel P is detected not as the pixel N of the characteristic point but as the pixel M near the characteristic point. Therefore, the pixel P is marked with a mark indicating that it is in the region of the pixel M of the characteristic point.

FIG. 8 shows a characteristic point symbol in which each arbitrary pixel belongs to a plurality of arbitrary pixels other than the characteristic point pixel.
FIG. 7 shows a state filled in according to the procedure as described above with reference to FIG. The dotted line connecting the above-mentioned feature point symbols is the boundary of the influence region of the above-mentioned feature points. Now, focusing on Δ1, which is the boundary of the three regions, for example, pixels Nj, Nk, and Ok are adjacent to this Δ1. Therefore, the pixel N
j, Nk, Ok are regarded as the three neighboring feature points, the triangles Nj, Nk, Ok are determined as the luminance interpolation plane, and the luminance values of the pixels existing inside the triangles Nj, Nk, Ok are It is filled with the brightness value of the brightness interpolation plane. Similarly, when attention is paid to Δ2 which is the boundary of another three regions, the pixels Nj, Oj and Ok are adjacent to this Δ2. Therefore, the pixels Nj, Oj, Ok are regarded as the three neighboring feature points, the triangles Nj, Oj, Ok are determined as the brightness interpolation plane, and the brightness of the pixels existing inside the triangles Nj, Oj, Ok is described. The value is assigned by the brightness value of the brightness interpolation plane described above.

By the procedure as described above, the pixel area affected by one feature point is searched by the extension plane (extended solid in the case of a three-dimensional object), and the pixel area affected by the one feature point is searched. By detecting three feature points in the vicinity (4 feature points in the case of a three-dimensional object) from the boundary, the brightness function on the decompression side is determined from a small number of feature point information, and each brightness of the untransmitted image information is determined. The value is applied by the above-mentioned interpolation value to reproduce the content of the entire image. It should be noted that circles, squares, rectangular parallelepipeds, rhombuses, and the like may be used as the expansion surface in the case of a two-dimensional object, and spheres, spheroids, and bottom surfaces are symmetrical as the expansion solid in the case of a three-dimensional object. Two cones connected as planes, two pyramids connected with the bottom as a plane of symmetry, and the like may be used.

By the procedure as described above, the brightness function on the decompression side is determined based on a small number of feature point information, and each brightness value in the untransmitted pixel information is interpolated by the brightness value on the brightness interpolation plane. The content of the entire image will be reproduced. As is clear from the above description, what is shared between the compression side and the decompression side is the brightness function of the screen, and if the position information of the feature points is indicated by a relative value for the entire screen, The number of pixels on the side does not have to match. That is, the characteristic information of the image extracted by the multidimensional color image compression method of the present invention is redrawn according to the number of pixels (resolution scale) on the decompression side.

In the multi-dimensional color image compression system of the present invention shown in FIG. 1, an image signal source (for example, a TV camera) 1 generates a video signal with a luminance, a hue, and a saturation in accordance with a predetermined television system to generate an analog signal. It is supplied to the digital converter 2. As the above-mentioned image signal source 1, any configuration mode can be used as long as it is a video signal generator capable of generating image information to be compressed and expanded by the multidimensional color image compression system of the present invention. However, in the following description, it is assumed that the image signal source 1 has a configuration capable of generating a color video signal of a moving image. The color video signal of the moving image including the brightness and the hue generated from the image signal source 1 is converted into a digital signal by the analog-digital converter 2. The analog-to-digital converter 2 decomposes the video signal for each image into a predetermined number of pixels in each of the horizontal and vertical directions of the image (for example, 512 pixels in the horizontal direction of the image and 480 pixels in the vertical direction of the image). A digital signal having a predetermined number of bits (for example, 8 bits) is set for each pixel in the state where
It is supplied to the write control unit 3 which controls writing to a and 4b.

The image memories 4a and 4b are image memories used for extracting contour lines of specific luminance and hue in an image and performing an operation for determining a characteristic point. The data writing operation and the data reading operation performed at the time of determining the characteristic point are sequentially and alternately performed in parallel in the two image memories. The outputs from the image memories 4a and 4b are
Contour tracers 5a and 5b (Detailed configuration examples are shown in FIGS. 3 and 2).
Supplied). Then, the contour line tracers 5a and 5b described above perform the contour line extraction from the image information and the feature point determination operation. The area indicated by the hue contour line of the contour tracer 5a is the hue of the human skin color described above (in the color space, the hue of the red primary color in the three primary colors of additive mixture is 0 degrees, and the hue of the blue primary color is 120 degrees. , When the hue of the primary color of green is represented as 240 degrees,
When the luminance value generators 18 to 20 in the contour tracer shown in FIG. 2 for extracting the luminance contour lines inside the area are other hue regions, It is possible to generate a high-density luminance value in a finer step than that of, and to extract a high-density contour line.
Then, the information on the characteristic points of the image which is the object of the determined image information compression is given to the encoding and transmitting circuit 6.

In the above-mentioned encoding / sending circuit 6, the information is converted into a code that can be efficiently transmitted, for example, a known code such as Huffman code and is transmitted to the receiving / decoding circuit 8 via the transmission line 7. The receiving / decoding circuit 8 decodes the coded signal transmitted thereto and supplies the decoded signal to the luminance function reproducing device 9 (a detailed configuration example is shown in FIG. 4). Needless to say, when the transmission line 7 is used as a recording medium, a recording circuit and a reproducing circuit are used as the encoding and sending circuit 6 and the receiving and decoding circuit 8, respectively. The luminance function reproducing device 9 restores the uncompressed two-dimensional luminance function using the feature point information supplied thereto. Then, the restored luminance value of each pixel forming the two-dimensional luminance function is read as a time-series signal in the form of an analog signal and supplied to the drive circuit 10. The drive circuit 1
At 0, a video signal of an image to be displayed by the monitor receiver 11 is generated and given to the monitor receiver 11.

The two contour tracers 5a and 5b shown in FIG. 1 are the contour tracers 5b of the luminance function after the contours of the color correlation number are extracted by the contour tracer 5a (see FIG. 3) as described above. Although used for extraction in FIG. 2 (see FIG. 2), the contour tracer 5a extracts the contours of the color correlation number and the contour tracer 5
Since the contour line extraction processing operation of the luminance function by b is the same, the contour line extraction processing operation of the color correlation number and the contour line extraction processing operation of the luminance function sequentially perform the same contour line tracer in a time division manner. It may be configured to be executed by utilizing it. However, as in the configuration example illustrated in FIG. 1, when the contour line extraction processing operation of the color correlation number and the contour line extraction processing operation of the luminance function are performed by individual contour line tracers, the signal processing operation is performed. It is possible to speed up. The contour line extraction of the color correlation number by the contour tracer 5a whose concrete configuration example is shown in FIG. 3 and the extraction of the contour line of the brightness function by the contour tracer 5b whose concrete configuration example is shown in FIG. Since the processing operation is the same, the contour tracer 5b will be used as a representative of the contour tracer, and the contour tracer 5b shown in FIG. By doing so, the contents of the concrete contour line extraction processing operation in the above-mentioned two contour line tracers 5a and 5b will be clarified. 2 and 3
In the contour line tracers 5a and 5b illustrated in FIG. 3, the illustration of the components for bypassing the saturation information is omitted for simplicity of illustration.

In FIG. 2 showing a detailed configuration example of the contour line tracer 5b, the contour line tracer 5b and two image memories 4a and 4b (in FIG. 2, 2 shown in FIG. 1) are shown.
It is shown that a plurality of signal transmission paths are provided between each of the image memories 4a and 4b. The contour tracer 5b has a plurality (n) of address generators 12 to 14 and a brightness determiner 15 respectively.
-17, brightness value generators 18-20, shift register 21
23, feature point determiners 24 to 26, and one alignment circuit 2
7 are provided. Address generator 12-
At 14, address data designating addresses of the image memories 4a and 4b are generated, and the generated address data are given to the image memories 4a and 4b.

The brightness value data read from the storage area of the address designated by the address data generated from the address generator 12 (13, ... 14) and the brightness value generating section 18 (19, ... 20) And the brightness determining device 15 (16, ... 1)
In 7), the above two data are compared and the comparison result is given to the address generator (13, ... 14). The step of the brightness value generated by the brightness value generation unit 18 (19 ... 20) is the hue of the human skin color (the red primary color of the three additive primary colors in the color space) from the output result of the hue contour tracer 5a. The hue is 0 degree and the hue of the blue primary color is 120
When the hue of the primary color of green is 240 degrees, the density is increased when the hue is in the range of 10 degrees to 30 degrees. As a result, the address generator 12 generates new address data (including the middle of the pixel address) corresponding to M1 in FIG. 5 (b) (the other address generators 13 ... It goes without saying that data is generated)
a and 4b. In the image memories 4a and 4b, the brightness value data read from the storage area at the address designated by the address data given as described above is stored in
Brightness determiner 15 (16, ... 17) of the contour tracer 5b
Supply to.

In the brightness determiner 15 (16, ... 17), the brightness value data read from the image memories 4a, 4b and the specific brightness value generated by the brightness value generating section 18 (19, ... 20) are stored. The data is compared and the comparison result is given to the address generator (13, ... 14). Hereinafter, in the same manner, the above-described operation is sequentially performed, so that the brightness value substantially equal to the brightness value indicated by the data of the specific brightness value generated by the brightness value generation unit 18 (19, ... 20). Are sequentially generated. When each of the above address values indicates the middle of the pixel address values as in the example illustrated in FIG. 5, it is replaced with the neighboring pixel address values, and the above operation is sequentially performed. .

As described above, the address generator 12 (13,
The address data sequentially generated from (14) is also given to the shift register 21 (22, ... 23), and the parallel output of the shift register 21 (22, ... 23) is supplied. In the determiner 24 (25, ... 26), one new address is input while the shift register 21 (22, ... 23) described above stores address data of a specific number of pixels (for example, 10 pixels). Each time one oldest address disappears, a pixel serving as a feature point is detected from the address group according to the above-described criterion for determining the feature point.
Then, the address value and the brightness value of the pixel that is the characteristic point are supplied to the alignment circuit 27. In the alignment circuit 27, as described above, the address value and the brightness value of the feature point group indicating different contour values output from the plurality of feature point determiners 24, 25, ... An alignment operation is performed for sorting and sending. The output from the alignment circuit 27 is
It is supplied to the encoding transmission circuit 6 as the output of the contour line tracer 5b. As the image memories 4a and 4b, those having a multiport structure are used because the signal processing in the contour line tracers 5a and 5b is adapted to trace a plurality of contour lines in parallel.

In terms of time, the contour tracer 5a,
When the scanning of one screen by 5b (one field period for the image in which the interlaced scanning method is adopted as the scanning standard) is completed, the feature point extraction operation for the one screen is completed. And contour line tracer 5
In the encoding / transmission circuit 6 to which the output of b is supplied, as described above, the information is converted into a code that can be efficiently transmitted, for example, a known code such as a Huffman code, and is transmitted to the reception / decoding circuit 8 via the transmission line 7. To transmit. Execution of the feature point extraction operation for the next screen of the screen after the feature point extraction operation is completed, the time required for encoding to convert to a known code such as the above-mentioned Huffman code, Transmission line 7
During the time required for transmission to the receiving and decoding circuit 8 via. Note that if the sum of the encoding time and the transmission time may exceed the scanning time of one screen due to a large number of detected feature points in one screen, for example, It is a desirable embodiment to automatically adjust the above-mentioned threshold angle or threshold distance used for the point detection operation so that the number of characteristic points is limited.

The receiving / decoding circuit 8 decodes the coded signal transmitted thereto and supplies it to the luminance function reproducing device 9. 4, which shows a detailed configuration example of the luminance function reproducing device 9 shown as a block 9 in FIG.
28 is a switcher, 29 and 30 are image memories, 31 is a search address generator, 32 is an interpolation surface determination circuit, 33 is an interpolation value writing circuit, 34 is a TV reading circuit, 35 is a digital-analog converter, and a brightness function In the playback device 9,
Using the feature point information supplied from the receiving circuit 8, the operation of restoring the two-dimensional luminance function before compression is performed. Information of feature points for the luminance function reproducing device 9 shown in FIG.
That is, when the address value and the brightness value of the feature point are supplied as input information, the information is supplied to either one of the two image memories 29 and 30 via the switch 28,
Remembered in it. The two image memories 29, 30
The reason why it is used by switching is that one of the image memories used for the expansion operation is
This is because by separately providing an image memory used for reading and displaying the image information of the decompressed image memory, simultaneous operations are possible and the entire decompression work is speeded up.

The search address generator 31 generates addresses of pixels other than the pixels of the above-mentioned feature points written in the image memory 29 (or 30), and the pixel of the feature point closest to the above-mentioned pixel is displayed. 7 is searched by the method as described above, and a mark indicating that the pixel is located in the region of the pixel of the closest feature point (attribute region of the pixel of the feature point) is entered as shown in FIG. Further, the interpolation surface determination circuit 32 searches for an area boundary by reading out the attribute area in the image memory 29 (or 30) described above, and 3 is used for interpolation determination from the names of the added feature points of the adjacent areas.
Extract individual feature points. Further, the interpolation value writing circuit 33 calculates the interpolation value of the internal area of the interpolated triangular surface determined in the image memory 29 (or 30), and the pixels other than the feature points existing in the internal area are calculated. Luminance value and chromaticity value of
The image memory 2 described above so as to be filled with the interpolation values described above.
Write to 9 (or 30).

Further, the TV reading circuit 34 uses the information of pixels stored in the other image memory 30 (or 29), which is currently being worked, as the predetermined standard system. The time-sequential image signal is read out. The time-sequential image signal is converted into an image signal in the form of an analog signal by the digital-analog converter 35, and then supplied to the drive circuit 10. The drive circuit 10 adds a blanking signal or a synchronizing signal to the image signal and supplies it to a monitor TV receiver (see FIG. 1) to display a reproduced image on the display surface of the monitor TV receiver. Display it.

The description so far has been made on the case where the multi-dimensional color image compression method of the present invention is applied to the reproduction of the two-dimensional luminance function, the color correlation number and the saturation function. However, the multi-dimensional color of the present invention is applied. The image compression method can be well applied to the three-dimensional image compression method in which the feature points are extracted over a plurality of screens arranged on the time axis. Next, referring to FIG. 9, the multi-dimensional color image compression of the present invention will be described. An example in which the method is applied to a three-dimensional image compression method in which feature points are extracted over a plurality of screens arranged on the time axis will be described. 9 is 3
An example of a case in which the luminance information of pixels other than the pixels of the feature points is determined by the interpolated solid determined by the feature points detected from the image information distributed in a dimension will be described. In FIG. 9, images P1 to P4 show still image groups arranged on the time axis, and the image information as a whole has a three-dimensional distribution.

Still images P1 to P4 shown in FIG.
With respect to the two-dimensional luminance function in, the equal luminance line is obtained by tracing with equal luminance, but the equal luminance surface is obtained by expanding it in three dimensions. Then, in FIG. 9, the characteristic points of the image are the positive and negative maximum points of the curvature of the equal-luminance surface, or the points where the error from the approximate plane of the equal-luminance surface exceeds a predetermined threshold value. M1 to M4. Then, when it is determined that the above-mentioned feature points are three-dimensionally close,
Interpolated solids M1, M2, M3, M4 are defined. The projection of the still image P2 of the interpolated solids M1, M2, M3, M4 is a triangle M2, N1, N2, and the projection of the still image P3 of the interpolated solids M1, M2, M3, M4 is The triangles M3, N3, N4. The brightness value of each vertex of each triangle can be determined by performing an interpolation calculation from each brightness value of each vertex of each of the interpolated solids M1, M2, M3, and M4. It is possible to determine the luminance value of each pixel existing inside each triangle by the same method as that for determining the untransmitted pixel luminance by the interpolation plane.

Further, as shown in FIG. 9, without detecting the three-dimensional feature points as described above, for example, the two-dimensional feature points are detected for the still image P1, and as a result, the feature points O1 and O2 are obtained. However, as a result of performing the two-dimensional feature point detection on the still images P4 separated by a certain time,
If the characteristic points O3 and O4 are obtained and they are three-dimensional neighboring points, the interpolated solids O1, O2, O3, O4 are defined, and the interpolated solids O1, O2, O3, O4 are defined. The projection by the still image P2 is quadrangle I1, I2, I3, I4,
In addition, the still image P of the interpolated solids O1, O2, O3, O4 described above.
The mapping by 3 becomes squares I5, I6, I7, and I8. In this case, the polygon is different from the triangle in the case of the above-described interpolation plane, but since a general polygon can be divided into triangles, a single interpolation plane can be defined even in this case, and the two-dimensional Similar to the above method, the brightness value of each pixel existing inside each triangle can be determined by the same method as the case of determining the untransmitted pixel brightness by the interpolation plane for the two-dimensional image information described above.

[0045]

As is clear from the above description, the multi-dimensional color image compression method of the present invention enables the image processing regardless of whether the pixel density of the image targeted for image information processing is high or low. Extracting only the characteristic points of the image data to obtain the image data in which the image information is compressed, and when decompressing, the pixel restoration is not performed from the above-mentioned image data, but a new image is formed on another pixel density surface. In order to be able to draw, the two-dimensionally distributed luminance information, the two-dimensionally distributed luminance information, and the three-dimensionally distributed image information including the time axis in the image information of chromaticity information Luminance function, the positive and negative maximum points of the curvature of the contour line of the color correlation number, or the positive and negative maximum points of the curvature of the contour surface as the characteristic points, or the straight line that approximates the contour line of the luminance and color correlation numbers and the above-mentioned luminance and hue Error from the contour of the function, Or, the error between the plane approximated by the plane approximated by the contour plane of luminance and color correlation number and the contour plane of the luminance and color correlation number mentioned above is the point where the threshold value exceeds a predetermined threshold value, and the above-mentioned image When the position, the luminance value, and the chromaticity value of the feature point are transmitted and recorded, and the position, the luminance value, and the chromaticity value of the feature point are used for image restoration, they are determined by a plurality of neighboring feature points at the time of expansion. The luminance and chromaticity information of pixels other than the feature points are determined by the interpolated surface or the interpolated solid that is generated,
From the three-dimensionally distributed image information that includes the time axis in the image information of the three-dimensionally distributed luminance and chromaticity information,
For multiple sets of dimensionally distributed luminance and chromaticity information,
The brightness of each of the groups, the positive and negative maximum values of the curvature of the contour line of the color correlation number, or the brightness of each of the groups, the above-mentioned luminance and a straight line linear approximation of the contour line of the color correlation number, the error between the contour line of the color correlation number, When a point exceeding a predetermined threshold value is used as an image feature point, and when the position of the image feature point, the luminance value, and the chromaticity value are used for transmission, recording, and image restoration, a plurality of neighboring pixels are used for expansion. The multi-dimensional color image of the present invention can be easily expanded by determining the luminance and chromaticity information of pixels other than the feature points by the interpolated solid determined by the feature points. In the image compression method, regardless of whether the pixel density of the image that is the object of image information processing is high or low, only the characteristic points of the image are extracted to obtain the image data in which the image information is compressed. Image day An image of two-dimensional luminance, three-dimensional luminance including chromaticity information and time axis, chromaticity information, etc. so that a new image can be drawn on another pixel density plane instead of performing pixel restoration from When the characteristic points of the image described above for information are obtained and the image obtained by using the positions of the characteristic points of the image and the luminance value and the chromaticity value are restored, Za = Σ (zi / ri) / Σ (1 / ri) (A) where Za is the restored luminance value zi is the luminance value of each feature point ri is the distance element from each feature point to the restoration point i Feature point number (A) I made it possible to easily expand the multi-dimensional color image using
In the method of the present invention, high compression is performed according to the amount of information of an image without being related to physical conditions such as a difference in the number of constituent pixels between the original image to be compressed image information and the reproduced image obtained by decompression. Can be easily realized.

[Brief description of drawings]

FIG. 1 is a block diagram of a multi-dimensional color image compression / expansion system configured to include the multi-dimensional color image compression system of the present invention.

FIG. 2 is a block diagram showing a configuration example of a contour tracer used in the multi-dimensional color image compression system of the present invention.

FIG. 3 is a block diagram showing a configuration example of a contour tracer used in the multi-dimensional color image compression system of the present invention.

FIG. 4 is a block diagram showing a configuration example of a luminance function reproducing device.

FIG. 5 is a diagram for explaining a method of detecting a luminance contour line and a method of determining a feature point.

FIG. 6 is a diagram used for explaining a configuration principle and an operation principle of a luminance function reproducing device.

FIG. 7 is a diagram used for description of luminance function reproduction.

FIG. 8 is a diagram used for explaining a surface interpolation operation for reproducing a luminance function.

FIG. 9 is a diagram used for explaining a case where the luminance information of pixels other than the feature points is determined by the interpolated solid determined by the feature points detected from the image information distributed three-dimensionally.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Signal source of the image signal used as the object of multidimensional color image compression / expansion 2 ... Analog-digital converter 3 ... Writing control part 4a, 4b, 29, 30 ... Image memory 5a, 5b
... Contour tracer, 6 ... Encoding transmission circuit, 7 ... Transmission line
(Or recording medium), 8 ... Reception decoding circuit (or reproduction decoding circuit), 9 ... Luminance function reproducing device, 10 ... Driving circuit, 1
1 ... Monitor receiver, 12-14 ... Address generator, 15
-17 ... Luminance determiner, 18-20 ... Luminance value generator, 24
... 26 ... Feature point determiner, 21-23 shift register, 2
7 ... Alignment circuit, 31 ... Search address generator, 32 ... Interpolation plane determination circuit, 33 ... Interpolation value writing circuit, 34 ... TV reading circuit, 35 ... Digital-analog converter,

Claims (11)

[Claims] 1. Regarding image information including a two-dimensionally distributed color in a still image, a characteristic point of an equihue line of a color correlation number and an isoluminance line of a brightness function in a region included in the above-mentioned equihue line. The multi-dimensional color image compression method used for transmitting, recording, and image restoration of the position of the feature point, the luminance value, and the chromaticity value of the image information described above. 2. The two-dimensionally distributed image information from among the three-dimensionally distributed image information in which a time axis is added to the image information including the two-dimensionally distributed color in a still image. For a plurality of sets of, to obtain the characteristic points of the equal hue line of the color correlation number of each set, and the characteristic points of the equal luminance line of the luminance function in the area included in the equal hue line, The position of the characteristic point of the image information, the luminance value, and the chromaticity value are transmitted,
A multi-dimensional color image compression method used for recording and image restoration. 3. The multi-dimensional color image compression method according to claim 1, wherein the positive and negative maximum points of curvature of contour lines of the target two-dimensional function are used as feature points. 4. A feature point is that an error between a straight line obtained by linearly approximating a contour line of a target two-dimensional function and a contour line of the target two-dimensional function exceeds a predetermined threshold value. 3. The multi-dimensional color image compression method according to claim 1 or 2. 5. With respect to the three-dimensionally distributed image information in which a time axis is added to image information including a two-dimensionally distributed color in a still image, characteristic points of an equal hue plane of a color correlation number, and The characteristic points of the uniform luminance surface of the luminance function in the area included in the uniform hue plane are obtained, and the positions of the characteristic points of the image information, the luminance value, and the chromaticity value are transmitted, recorded, and image restored. Multi-dimensional color image compression method used. 6. The multi-dimensional color image compression system according to claim 5, wherein the positive and negative maximum points of the curvature of the contour surface of the target three-dimensional function are used as feature points. 7. An error between a plane that is a plane approximation of the contour surface of the target three-dimensional function and the contour surface of the target three-dimensional function exceeds a predetermined threshold value. The multi-dimensional color image compression system according to claim 5, wherein points are used as feature points. 8. An image compression method used for transmitting, recording, and restoring image information of isoluminance lines of a luminance function in an equihue plane for image information including a two-dimensionally distributed color in a still image, wherein the hue is Separately, a multidimensional color image compression method in which the isoluminance linear density is changed. 9. The two-dimensionally distributed image information among the three-dimensionally distributed image information in which a time axis is added to the image information including the two-dimensionally distributed color in a still image. In the image compression method used for transmitting, recording, and image restoration of the information about the equiluminance lines of the luminance function in the equal hue plane of each of the above sets, the equal luminance line density is changed for each hue. Multi-dimensional color image compression method. 10. The image information which is three-dimensionally distributed by adding the time axis to the image information including the two-dimensionally distributed color in the still image, is the same luminance surface of the luminance function in the equal hue solid. An image compression method used for information transmission, recording, and image restoration, which is a multidimensional color image compression method in which the uniform luminance surface density is changed for each hue. 11. The hue for changing the equal-luminance linear density and the equal-luminance area density is the hue of human skin color, and the change is performed in the direction of increasing the density. 10. A multi-dimensional color image compression method according to any one of 10.