KR20010036671A - Block-based Image Histogram Generation Method - Google Patents

Abstract

PURPOSE: A method of generating block-based image histogram is provided to be capable of extracting specific information from a pixel group of a block unit by use of a block of plural pixels by a basic unit of generating histogram. CONSTITUTION: A coefficient 'k' is set to '0', and pixels of an image are grouped to set a block size of a level '0'(S101). Specific information of the image is extracted by a block unit with regard to all blocks of the level '0' to update a corresponding histogram(S102-S105). After extracting the specific information with regard to all blocks of the level '0', the coefficient 'k' indicating the level is increased by '1'(S106). Specific information of a level '1' is generated by merging specific information of the level '0'(S107), updating a corresponding histogram(S108). Operations of extracting the specific information with regard to all blocks of the level '1' and updating a corresponding histogram are repeated(S109,S110). If the operation of generating specific information with regard to all levels is not completed(S111), the procedure goes to the step(S106) of increasing the coefficient. If the operation of generating specific information with regard to all levels is completed(S111), the procedure is ended.

Description

Block-based Image Histogram Generation Method

The present invention relates to a method for generating normalized histograms from compressed data bitstreams or uncompressed image data conforming to the standards of JPEG, MPEG-1, 2, and in particular, using color-by-block linear quantization, color, brightness, and edges. A method of generating a histogram with components is provided.

JPEG is an international standard for storage and transmission in still images, and MPEG-1 and 2 are very useful. It is necessary to extract feature information for each image for applications such as key frame extraction, image retrieval, and browsing for the compressed image information.

As described above, in order to extract feature information from an image, a brightness or color histogram representing a relative occurrence frequency of brightness or color (red, green, and blue) in the image is frequently used. In particular, a method of comparing histograms for retrieval of still images or digital video that has been digitized and stored has been recently proposed. As such, histograms are widely used for image retrieval and scene change detection. Therefore, improvements on histograms are required. In other words, by developing a discrete histogram of discrete quantization and color only, a composite histogram of color, edge, and brightness using linear update or soft decision is adopted. There is a need for more efficient and faithful representation.

A description of a conventional histogram generation method is as follows.

U.S. Patent No. 5,805,733, entitled 'Katherine et al.', Entitled 'Method and system for detecting scenes and summarizing video sequences' describes the detection of scene transitions using color histograms and edge maps. This has the effect of extracting color information in consideration of human vision, but does not use brightness. 'M.J.Swain and one other person' said Int. 'Color Indexing', a paper published in J. of Computer Vision, describes the indexing by measuring the similarity of images using histogram intersection. However, this also has a problem that accuracy is lowered because it does not use the brightness and edge information. In addition, since these histograms generate histograms using discrete quantization techniques, a relatively large number of histogram bins are required to achieve the same performance. Therefore, storage and similarity measurement are inefficient.

In addition, conventionally, since feature extraction is performed on a pixel basis when generating a histogram, there is a problem that only limited feature information can be generated.

Accordingly, the present invention has been made to solve the above problems of the prior art, by using a block consisting of a plurality of pixels as a basic unit of histogram generation, the brightness, color, A block-based image histogram generation method for extracting feature information such as edges is provided.

In addition, the present invention is to separate the meaning of each bin of the histogram in the image histogram, and to generate a histogram that more effectively reflects the contents of the image.

1 is a flowchart illustrating a method of generating a block-based image histogram according to an embodiment of the present invention;

2 is a flowchart illustrating a block-based image histogram bin updating method according to an embodiment of the present invention;

3 is a diagram illustrating brightness representative values and types of edges of subblocks in a block;

4 illustrates a block filter for edge detection;

5 is a flowchart illustrating an edge detection method;

6 is a diagram illustrating a process of generating a block of a higher level by merging a plurality of blocks of a lower level;

7 is a diagram illustrating an example of a histogram semantic;

8 is a diagram illustrating a conventional histogram updating method without considering linear weights;

9 is a diagram illustrating a histogram updating method according to an embodiment of the present invention considering linear weights;

FIG. 10 is a diagram illustrating linear weights between chromatic bins and bins and linear weights between chromatic and achromatic colors; FIG.

11 is a diagram illustrating an example of a composite histogram constructed by using brightness characteristic information and edge characteristic information;

12 is a composite histogram constructed by using feature information in a global area, feature information in a semi-global area, and feature information in a local area;

FIG. 13 is a diagram illustrating an example of the complex histogram of FIG. 12.

According to the present invention for achieving the above object, in the method for generating an image histogram using the color information and the brightness information of the image data, the feature information of the image in units of blocks that group the pixels of the image into a group A histogram is generated by extracting the image, changing the size of the block, and extracting feature information of the image in units of the changed size of the block to generate a histogram.

Preferably, the first step of initializing the variable k and grouping the pixels of the image into a group to set the block size of the level k, extracting the feature information of the image in units of blocks for all the blocks of the level k A second step of updating the histogram, and combining the blocks of the level k to create a level k + 1 block, merging the image characteristic information of the level k, and performing image-wise block-by-block for all blocks of the level k + 1; And generating a feature information of the third step and updating the corresponding histogram, and repeating the third step for every level while increasing k by one.

Here, the feature information of the image includes brightness, edge, and color information.

More specifically, the step of extracting the edge feature information of the image in the second step, the first sub-step dividing the block into two horizontally and vertically divided into four sub-blocks, the brightness of each of the sub-blocks A second sub-step of obtaining representative values, and if the brightness difference between the adjacent sub-blocks exceeds the threshold Te, it is determined that an edge exists between the sub-blocks, and the type of edge (horizontal edge, vertical edge, mixed) A third substep of detecting an edge). If the brightness difference between all the subblocks does not exceed the threshold Te, the block further includes a fourth substep of determining that the monotone block has no edge.

More preferably, the step of extracting the edge feature information of the image in the second step, the first sub-step of dividing the block into two horizontally and vertically divided into four sub-blocks, the brightness of each of the sub-blocks A second substep of obtaining representative values, and convoluting the brightness representative values and the filter coefficient values of the subblocks with a zero degree edge value (edge0), a 45 degree edge value (edge45), and a 90 degree edge. A third substep of obtaining the value edge90, the 135 degree edge value edge135, and the mixed edge value complex_edge, respectively, the 0 degree edge value edge0, the 45 degree edge value edge45, and the 90 degree value A fourth sub-step of comparing the edge value edge90, the 135-degree edge value edge135, and the maximum edge value among the mixed edge value complex_edge and the threshold value Th_sub_differece, and the maximum edge value in the fourth sub-step. If it is larger than this threshold value, the fifth book determines the edge of the maximum value as the representative edge of the block. And a step. Further preferably, the fourth sub-step further includes a sixth sub-step of determining the block as a single tone block if the maximum edge value is smaller than a threshold.

Even more preferably, the third sub-step may include applying the brightness representative values and the filter coefficient values of the sub-block to Equation 1, and the 0 degree edge value, the 45 degree edge value, and the 45 degree edge value, The 90-degree edge value edge 90, the 135-degree edge value edge 135, and the mixed edge value complex_edge are calculated.

In this case, the brightness representative value of the subblock uses an average value of pixels or a median value of pixels.

More preferably, the extracting the color and brightness characteristic information of the image in the second step may include a first sub-step of obtaining a pseudo color tone (ph) and a pseudo color density (ps) using color difference information of a block; A second substep of calculating a linear weight (α) between the colored and empty bins and a linear weight (β) between the colored and achromatic colors using the pseudo color tone and the pseudo color concentration, using the pseudo color concentration value; A third sub-step of determining whether the color of the block is pure color or a dark chromatic color including achromatic color; if the color of the block is pure color, the increase of the corresponding color histogram bin is calculated by using the linear weight α A fourth sub-step of updating a color histogram, and if the color of the block is a dark chromatic color, using the linear weight α and the linear weight β, a corresponding color histogram bin And a fifth substep of calculating the increment of the brightness histogram bin and the increment of the brightness histogram bin to update the corresponding color histogram and the brightness histogram.

Preferably, at least two or more features of the image are represented in one composite histogram.

According to the present invention, there is provided a computer-readable recording medium having recorded thereon a program for realizing the above-described method for generating a block-based image histogram.

Hereinafter, a method of generating a block-based image histogram according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

The characteristics of the present invention can be seen in five large. First, feature information is extracted by grouping a plurality of pixels into one block and using this as a basic unit. Second, the edge and texture information are included in the feature information extracted in block units. Third, the size of the block can be modified according to the size of the image so that quantitative feature information can be extracted regardless of the size of the entire image. Fourth, in calculating the increase of the histogram, the linear weight is applied to reduce the quantization error. Fifth, each histogram is generated for an entire region, a partial entire region, and a local region for one image.

1 is a flowchart illustrating a method of generating a block-based image histogram according to an embodiment of the present invention.

Referring to FIG. 1, first, a variable k is initialized to 0 and a block size of level 0 is set by grouping pixels of the image into a group (S101). The feature histogram is updated for each block of the level 0 in units of blocks (S102, S103, S104, and S105). That is, extracting feature information in a block from the first block (S102) and updating the corresponding histogram (S103) is performed for all blocks (S104 and S105). The feature information in the block includes brightness, color, and edge information, and a detailed description of extracting information will be described later.

Next, after feature information is extracted for all blocks of level 0, the variable k representing the level is increased by 1 (S106), and then the feature information of level 0 is merged to generate feature information of level 1 (S107). The histogram is updated (S108). In this case, the block of level 1 may be formed by grouping the blocks of level 0, and the feature information of level 1 may be generated by merging the feature information of level 0. The operation of extracting feature information and updating a corresponding histogram for all blocks of level 1 is repeatedly performed (S109 and S110).

Next, if feature information generation for all levels is not finished (S111), the process proceeds to step S106 in which the variable representing the level is increased by one, and ends when feature information generation for all levels is finished.

First, the basic unit of the histogram, which is the first characteristic, is described.

In the conventional histogram generation method, the brightness or color feature information extracted for each pixel in the image is updated only for the corresponding histogram bin. However, in the present invention, as shown in FIG. 2, the histogram bin is updated in units of one large pixel (block) by grouping several pixels, that is, N × M pixel groups, into one block. That is, by updating the histogram bin in block units, the edge feature information having various resolutions as well as the brightness and color characteristics of the image are extracted. In addition, by changing the size of the block in proportion to the size of the image, it is possible to extract quantitative feature information regardless of the size of the entire image (scale invariant).

2 is a diagram illustrating an example of a histogram update for each block unit of an image. Feature information extracted for each block is used to update the histogram bin on a block basis.

2, the feature information extracted from the second feature block will be described.

As above, N × M pixels are grouped into blocks to form a block, and feature information is extracted in units of blocks. Therefore, various feature information such as edges and textures generated in the blocks can be obtained as well as brightness and color information. Can be.

The feature information extracted in units of blocks can be reflected in the composite histogram update. One block consists of four sub-blocks as shown in FIG. Each subblock is represented by the brightness representative value of the inner pixels. As a representative value of the sub block, an average value or a median value of pixels in the sub block is used. The edge of a block composed of such sub blocks can be detected in the following manner.

A first embodiment of the method for detecting the presence or absence of an edge of a block is as follows.

As shown in (a) of FIG. 3, a block is composed of four sub blocks, and the brightness representative values of each sub block are DC1, DC2, DC3, and DC4, respectively. When the threshold value of the brightness difference for distinguishing the edges between blocks is Te, when the difference in the brightness values between the blocks exceeds Te, the edges exist between the corresponding blocks, which are represented by a thick solid line. When the edge exists in the block, it corresponds to one of three types of (b), (c) and (d) of FIG. 3. That is, it is a vertical edge like (b), a horizontal edge like (c) or a mixed edge like (d). When the edge is determined in this manner, the corresponding histogram bin is incremented by one.

A second embodiment of the method for detecting the presence or absence of an edge of a block is as follows.

4 is a diagram illustrating five filters used when detecting the presence or absence of an edge of a block according to the present embodiment. (A) is a filter coefficient value of edgd 90, (b) is a filter coefficient value of edge 0, (c) is a filter coefficient value of edge 45, (d) is a filter coefficient value of edge 135, and (e ) Is the filter coefficient value of the complex_edge. 5 is a flowchart illustrating a method for detecting the presence of an edge of a block according to the second embodiment of the present invention.

Referring to FIG. 5, the 0 degree edge value edge0, the 45 degree edge value edge45, the 90 degree edge value edge90, the 135 degree edge value edge135, and the mixed edge value complex_edge are obtained, respectively (S501). . In this case, the edge values may be obtained by convolution of the five filters and the representative values of the respective sub-blocks shown in FIG.

Here, mean_sub_block (0) means DC1 of FIG. 3 (a), mean_sub_block (1) means DC2, mean_sub_block (2) means DC3, and mean_sub_block (3) means DC4, respectively. In addition, edge90_filter (i) to complex_edge_filter (i) mean filter coefficient values shown in FIG. 4. For example, edge45_filter (0) is , edge45_filter (1) and edge45_filter (2) are 0, edge45_filter (3) is to be.

As described above, the 0 degree edge value (edge0), the 45 degree edge value (edge45), the 90 degree edge value (edge90), the 135 degree edge value (edge135), and the mixed edge values (complex_edge) are calculated, and among these, It is determined whether the largest value exceeds the threshold Th_sub_differece (S502) to distinguish a single tone block from an edge block. That is, when the maximum edge value is larger than the threshold value, the maximum edge value is determined as the representative edge of the block (S503). However, if the maximum edge value is less than the threshold value, the block is determined as a single tone block (S504). As described above, when the edge type of the corresponding block is determined, the corresponding histogram bin is increased by one.

3. Variable size block of feature information extracted from blocks

Up to now, the distribution of brightness, color, edge, etc. of the entire image is extracted at a constant resolution, but in order to express the image more effectively, it is necessary to extract feature information having various resolutions. In the present invention, as shown in FIG. 6, an upper block (level 1) consisting of several basic blocks (level 0) is defined. That is, the blocks of the lower level k are collected and defined as the blocks of the upper level k + 1. In the figure, blocks of level k range from (0,0) to (5,5). When nine blocks of lower level k are grouped, blocks from (0,0) to (2,2) become B00 blocks of upper level k + 1, and blocks from (0,3) to (2,5) Blocks become B01 blocks of higher level k + 1, blocks from (3,0) to (5,2) become B10 blocks of higher level k + 1, and (3,3) to (5,5) The blocks up to become B11 blocks of higher level k + 1.

To represent the image more effectively, edge components are extracted for each level to generate bins of the histogram. The feature information of the upper level block can be obtained by a simple operation using the feature information of the lower level blocks. The upper block feature information in which the location information in the image is added may configure a histogram of a semantic reflecting the feature information for each location and resolution in the image as shown in FIG. 7. That is, the vertical edge histogram bin, the horizontal edge histogram bin, and the mixed edge histogram bin may be separately generated for each level.

4. linear update

In color, the conventional histogram bin updating method uses a binary determination method of incrementing the bins of the histogram by 1 with or without data in a given region as shown in FIG. In this case, data existing near the boundary between the bins reflects similar color features, but since they are differently assigned to the two bins, serious errors represented by different feature information are generated. In order to eliminate this error, conventionally, the result histogram is flattened, that is, low pass filtering. However, due to such flattening, there is a problem in that the feature information is lost so that the accurate feature information of the image cannot be reflected.

In order to look at the problems of the prior art more clearly with reference to Figure 8 as follows.

The portions ″ A ″ in FIGS. 8A and 8C are all orange, but (a) is orange near red and (c) is orange near yellow. These two colors are similar orange, but the orange of (a) is counted in the red histogram bin as shown in (c), and the orange of (c) is counted in the yellow histogram bin as shown in (d). As described above, information existing near the boundary between bins has similar characteristics, but since they are assigned to different bins, there is a problem of generating a serious error represented by different feature information.

In order to solve this problem, the present invention reduces the quantization error generated between bins by applying linear weights according to the distance between each bin.

That is, referring to Fig. 9, both the portion ″ A ″ shown in (a) and the portion ″ A ″ shown in (c) are orange and are located near the boundary between red (R) and yellow (Y). At this time, orange close to red (R) is given more weight to red (R) as shown in (b) and histogram counts, and orange close to yellow (Y) is yellow as shown in (d). (Y) is weighted more and is histogram counted. By applying linear weights according to the distance between each bin, the quantization error occurring between the bins is reduced. Therefore, there is an advantage that the histogram having a small number of bins can accurately reflect the characteristic information.

An example of the application of such a linear update is as follows.

10 is divided into six types of magenta (M), blue (B), cyan (C), green (G), and yellow (Y), and achromatic color is divided into five types according to brightness. The figure shows in order to demonstrate the example which counts a bin. In this case, the histogram bins h (0) to h (5) are color histograms, and h (6) to h (10) are brightness histograms.

In the present invention, pseudo hue and pseudo saturation are calculated and used from the YCbCr color space. However, when using the color space represented by Hue and Saturation, the color tone and color concentration may be used as it is instead of pseudo hue and pseudo saturation.

In the YCbCr color space, Psuedo Hue and Psuedo Saturation are defined by Equation 2 according to Cb and Cr values.

The linear weight (α) between the color bins and the bins and the linear weight (β) between the achromatic and chromatic bins can be obtained as shown in Equation (3).

In this case, the color information using the color difference component is used only as color feature information irrespective of the brightness of the image. In the vicinity of the origin, which is the center of the color space, Cr and Cb have values near '0', so they are almost achromatic. In general, achromatic and chromatic colors are distinguished by using the distance between the color difference value and the origin as a threshold. In the case of a natural image, a lot of cases exist at this boundary, and an error occurring at this boundary occurs very seriously. In other words, the lighter colors are updated in the chromatic histogram bins, and the lighter colors are updated in the achromatic histogram bins. Occurs.

In the present invention, the update method considering the linear weight is used even near the boundary between the achromatic and chromatic colors.

If the starting threshold of the achromatic bin is Ts, since the pseudo color density (ps) exceeds Ts, since the color is pure color, the count of the color histogram is linear weight (α) between the bins as shown in FIG. Alone increases each bin (h (i)) in increments of the real value. This is expressed as in Equation 4.

Where i = 0,1,2,3,4,5

On the other hand, if the pseudo color density ps does not exceed Ts, that is, the color is close to achromatic color located near the origin in the color space, the color histogram is linear between the bins as shown in FIG. It increases by including the weight α and the linear weight β of the achromatic bin. This is expressed as in Equation 5.

Where i = 0,1,2,3,4,5

In addition, when the brightness is dark as described above, the brightness histogram is divided into five bins, and the brightness histogram bins are increased according to the condition shown in Equation 6 using the linear weight β.

Next, an example of updating a composite histogram composed of edges and brightness will be described. FIG. 11 is a diagram illustrating an example of applying brightness information of blocks of 0 levels and edge information of blocks of 0, 1, and 2 levels to histogram bin updating.

First, at the 0 level, the brightness histogram bins are divided into Q for the representative brightness value L of the block, and the division condition is expressed by Equation (7).

Then, the strengths of the directional edges obtained from the blocks at each level are respectively determined by the strength of the vertical edge (α _v (k)), the strength of the horizontal edge (α _h (k)), and the strength of the mixed edge (α _misc (k). ), The update of the histogram bin is as shown in Equation (8).

In the case of the complex histogram as described above, the zero-level brightness histogram bins provide h (0) to h (Q-1), and the zero-level to two-level edge histogram bins are h (Q) to h (Q + 8). ).

You can add a multi-resolution representation of brightness by adding a 1-level, 2-level, .., k-level brightness histogram.

5. Create histogram by region

As shown in FIG. 12, histograms are generated for the global region, the semi-global region, and the local region of the image by using the histogram generated in units of blocks. The histogram for each region generated as described above may be used as feature information capable of reflecting the overall feature of the image and local location information.

By using six brightness bins and five edge bins for the global region, eleven histogram bins can be generated as shown in FIG. 13. The 11 histogram bins are used to express the overall characteristic information of the image.

16 local regions are formed by dividing the entire region into four parts horizontally and vertically, and five edge bins are generated for each local region. As shown in FIG. 13, 80 histogram bins are generated.

Lastly, each partial area is accumulated for each type of edges horizontally and vertically to create four partial total areas for the horizontal and four partial full areas for the vertical. If five edge bins exist for each partial full region, 40 edge bins are generated as shown in FIG.

That is, in the composite histogram as shown in FIG. 13, brightness information of the entire area is updated from h (0) to h (5), and edge information of the entire area is updated from h (6) to h (10). The edge information for the local area is updated from h (11) to h (90) by 5 for each position, and the edge information for the partial entire area is 5 for each position from h (91) to h (130). Is updated.

When the histogram bins are determined in this manner, the positional information and the edge feature information are simultaneously expressed in the histogram of the partial entire region or the local region.

In this case, the composite histogram is normalized for each region, and the method is as follows.

First, the global region histogram bin (global_histogram_value [i]) is normalized as shown in Equation 9 by dividing by the number of blocks of the entire region existing in the image (global_number_block).

Second, the 80 histogram bins of the local area are normalized as shown in Equation 10 by dividing the number of blocks in the local area (local_global_number_block) with respect to the image of each local area.

Finally, 40 bins of the partial full region are normalized as shown in Equation 11 by dividing the number of blocks in the partial image horizontally or vertically by semi_global_number_block for each partial full region image.

The normalized nomalize_histogram [i] is expressed as shown in FIG. 13.

While the invention has been described above based on the preferred embodiments thereof, these embodiments are intended to illustrate rather than limit the invention. It will be apparent to those skilled in the art that various changes, modifications, or adjustments to the above embodiments can be made without departing from the spirit of the invention. Therefore, the protection scope of the present invention will be limited only by the appended claims, and should be construed as including all such changes, modifications or adjustments.

As described above, in the histogram generation method according to the present invention, since feature information is extracted in units of blocks composed of a plurality of pixels, edge information, which has not been conventionally extracted, may be extracted. In addition, since feature information such as brightness, color, and edge of various resolutions is extracted, feature information reflecting the content of an image can be extracted in more detail, and accurate comparison can be performed in similarity measurement or distance measurement.

In addition, by using the linear update method, there is an effect that the quantization error occurring at the boundary between bins can be significantly reduced. In addition, since the histogram bin is normalized and used, similar images can be searched regardless of the size of the image.

Claims (13)

In the method for generating an image histogram using color information and brightness information of image data, Extract the feature information of the image in units of blocks that group the pixels of the image into one group, update all histogram bins for each feature, change the size of the block, and extract the feature information of the image in units of the changed size block. And updating the partial histogram bins for each block, and combining the all histogram bins and the partial histogram bins to generate an image histogram. The method of claim 1, further comprising: initializing a variable k and grouping pixels of the image into a group to set a block size of level k; A second step of extracting feature information of an image in block units for all blocks of the level k and updating a corresponding histogram; Create a block of level k + 1 by combining the blocks of level k, merge the image feature information of level k, generate feature information of the image block by block for all blocks of level k + 1, and update the corresponding histogram. The third step, And a fourth step of repeating the third step for every level while increasing the k by one. The method of claim 2, wherein the feature information of the image includes brightness, edge, and color information. The method of claim 3, wherein the extracting edge feature information of the image in the second step comprises: A first substep of dividing the block into two horizontally and vertically to divide four subblocks; A second substep of obtaining brightness representative values of the respective subblocks, and When the brightness difference between the adjacent subblocks exceeds the threshold Te, it is determined that an edge exists between the corresponding subblocks, and a third substep of detecting a type of edge (horizontal edge, vertical edge, and mixed edge) is performed. Block-based image histogram generation method comprising the. The block-based image of claim 4, further comprising: a fourth sub-step of determining that the block is a monotone block having no edge if the brightness difference between all subblocks does not exceed a threshold Te. How to create a histogram. The method of claim 3, wherein the extracting edge feature information of the image in the second step comprises: A first substep of dividing the block into two horizontally and vertically to divide four subblocks; A second substep of obtaining brightness representative values of the respective subblocks, and The brightness representative values and filter coefficient values of the subblocks are convolutioned so that a zero degree edge value, a 45 degree edge value 45, an 90 degree edge value 90, and a 135 degree edge value are convolved. a third substep of obtaining edge135 and a mixed edge value complex_edge, respectively; The maximum edge value and the threshold value among the zero degree edge value edge0, the 45 degree edge value edge45, the 90 degree edge value edge90, the 135 degree edge value edge135, and the mixed edge value complex_edge A fourth sub-step of comparing (Th_sub_differece), and And a fifth substep of determining, as the representative edge of the block, the edge of the maximum value when the maximum edge value is larger than a threshold value in the fourth substep. The method of claim 6, And a sixth sub-step of determining the block as a single-tone block when the maximum edge value is less than a threshold value in the fourth sub-step. The method of claim 6, wherein the third sub-step, The brightness representative values and the filter coefficient values of the subblock are applied to the following equations for the 0 degree edge value, the 45 degree edge value, the 45 degree edge value, the 90 degree edge value, and the 135 degree edge value. and a step of obtaining the mixed edge value complex_edge (edge135). [expression] Here, edge90_filter (i) to complex_egde_filter (i) are coefficient values of the corresponding filter. The method of claim 4, wherein the brightness representative value of the subblocks is an average value of pixels. The method of claim 4, wherein the brightness representative value of the subblock is a median value of pixels. The method of claim 3, wherein the extracting color and brightness characteristic information of the image in the second step comprises: A first sub-step of obtaining a pseudo color tone (ph) and a pseudo color density (ps) using color difference information of a block, A second substep of calculating a linear weight? Between the colored and empty bins and a linear weight? Between the colored and achromatic colors using the pseudo color tone and the pseudo color concentration; A third sub-step of determining whether the color of the block is a pure color or a dark chromatic color including an achromatic color using the pseudo color concentration value; A fourth sub-step of updating the color histogram by calculating an increment of the corresponding color histogram bin by using the linear weight α when the color of the block is pure color; and If the color of the block is a dark chromatic color, the linear weight (α) and the linear weight (β) are used to calculate the increment of the corresponding color histogram bin and the brightness histogram bin to update the corresponding color histogram and the brightness histogram. Block-based image histogram generation method comprising the five sub-steps. The method of claim 1, wherein at least two or more pieces of characteristic information of the image are represented in one complex histogram. On your computer, A first step of initializing a variable k and grouping pixels of the image into a group to set a block size of level k, A second step of extracting brightness, color, and edge characteristic information of an image in units of blocks for all blocks of the level k and updating a corresponding histogram; Create a block of level k + 1 by combining the blocks of level k, merge the image feature information of level k, generate feature information of the image block by block for all blocks of level k + 1, and update the corresponding histogram. The third step, Recording a program for realizing a method of generating an image histogram using color information and brightness information of image data, including a fourth step of repeating the third step for all levels while increasing k by one. Computer-readable recording media.