JP2002142119A - Image compressing device and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the compressibility ratio of image data which are to be compressed reversibly and expressed in a color pallet mode. SOLUTION: A reconstituting section 10 reconstitutes the indexes of color pallets. In other words, the section 10 reconstitutes the indexes so as to get closer the indexes of continuous, picture elements in data processing order (for example, in the raster direction) in the image data so that the colors corresponding to the indexes may become suitable for 'data compression', though the colors are decided by appropriately assigning colors contained in the image data. When the picture elements which continue in the image data are taken into account, the possibility of higher-order bits becoming equal to each other when a bit plane is disassembled becomes larger and the compression efficiency in binary run length conversion is improved even when the colors largely change (for example, red and blue) by making the indexes of the largely changing colors to have close values.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a technique for compressing multi-valued image data represented by a color palette system.

[0002]

2. Description of the Related Art In recent years,
As information has been digitized, opportunities for handling image information have increased as a means of providing information to users. Image information is rich in instant comprehension, but the amount of data is enormous, so it is necessary to compress the data using a compression technique. For example, to process or display a color image in a computer or other digital image compression device, the color component values of each pixel in the image must be specified. For example, RG
In the B (red, green, blue) color space, a color image can be represented by designating a red intensity level, a green intensity level, and a blue intensity level for each pixel in the image. This is called a direct color method (FIG. 2 (a)
reference). It has been found that allocating 8 bits per color provides visually acceptable results. However, this requires 24 bits per pixel, which means that if high resolution is used frequently, a significant amount of data is required to represent a single color image. . Processing such large amounts of data requires both large amounts of memory and long processing times.

[0003] Conventional techniques have therefore been proposed for representing images using a smaller set of colors. Such color sets are conventionally known as color palettes and often include 256 (2 8) different colors. Each pixel in the color image can be represented by an 8-bit index of a color palette instead of 24-bit color information. This is called the color palette method (Fig. 2
(B)).

However, since there is no correlation between the index of the color palette and the color information, the compression efficiency is reduced from the viewpoint of data compression. That is, in the color pallet system, which is a limited color expression, it is essential to perform reversible compression, unlike the compression method for a normal direct color image. In the case of the direct color system, for example, even if the RGB values slightly change, the corresponding colors are not so different. This is because the data structure is not easily affected by lossy compression (see FIG. 3A). On the other hand, when irreversible compression is performed in the case of the color palette method, R
Because the index is changed instead of the GB value, the color becomes completely different. As described above, since there is no correlation between the index and the color information, it is natural that the colors corresponding to the adjacent indexes may be completely different (see FIG. 3B). Therefore, in the color palette method, it is impossible to use a method of simply discarding lower-order lower-order bits in order to reduce the data amount.

Accordingly, the present invention provides a compression processing technique capable of improving the compression ratio without deteriorating the image quality even if the image data is multi-valued image data expressed in a color palette system. Aim.

[0006]

The image data to be processed (compressed) by the image compression apparatus according to claim 1 is an index of a color palette. The color corresponding to the index is obtained by appropriately assigning the color included in the image data, and the color palette is reconstructed so as to be “suitable for data compression”. That is, the color palette is reconfigured so that the indices of consecutive pixels in the image data in the order of data processing are close to each other. Thereafter, the data is decomposed into bit planes, and the data of each plane is subjected to binary run-length conversion processing sequentially or in parallel, and a code is assigned to the converted data.

In other words, considering continuous pixels in image data, even if the color greatly changes, if the indexes of the greatly changing colors are set to be close to each other, the bit plane can be changed. There is a high possibility that the upper bits will be equal when decomposed.

For this reason, when binary run-length conversion processing is performed on the data of each plane obtained by decomposing a bit plane, the possibility that the same data continues is relatively high.
By performing the encoding, the compression efficiency is relatively improved. In the encoding, fixed-length encoding may be used, but if a variable-length code is assigned, further improvement in compression efficiency can be expected. As the variable length code, for example, a Huffman code or the like can be used.

If the indices are close to each other, the probability that the upper bits become equal when the bit plane is decomposed is large on the whole.
There are situations where this is not the case "locally". That is, even if the indices are close to each other, it is conceivable that the upper bits are hardly equal. For example, decimal 1
When bits 27 and 128 are expressed in bits, "0111"
1111 "and" 10000000 ". These are continuous as indices, but do not have exactly the same value in bit representation. On the other hand, when the decimal representation of 128 and 129 is represented by bits, "10000000" and "1
0000001 ", in which case the upper 7 bits are equal and only the least significant bit differs. Therefore, in the case of reconstructing the data so that the values are close to each other, more compression is achieved by allocating to an index such that the number of planes in which the data of the corresponding planes have the same value is relatively large. Efficiency is improved.

According to the above-described technical idea of the present invention, the point of improving the compression efficiency is how close the indices of consecutive pixels in the order of data processing in image data are. Needless to say, the compression ratio is further improved by devising as in claim 2, but here, in the sense of including such room for consideration,
Let us consider "how close the indexes are to each other".

[0011] For example, if the indices of the color pallet colors that are used are made continuous by separating the indices that are not used, the indices of the above-mentioned continuous pixels become closer to each other. In addition, the continuity of the data after the data conversion is improved, and the compression rate can be further improved.

Further, if the indices of the color pallet colors are assigned in order in correspondence with the order of the colors used in the image data in large amounts, the colors of the colors used frequently are Since the indices are close to each other and the probability that the values of the indices that are close to each other in the image data are continuous increases, a further improvement in the compression ratio can be expected.

Further, when a large number of colors are used in an image, a method according to claim 5 can be considered as to which index is assigned to which color. In other words, a pair of two colors is determined by a method of determining the largest color other than the focused color in the data before and after the focused color as the index before or next to the focused color, and determining the pair. For a color that does not exist, a pair is determined recursively by the above-described determination method, and the pair is arranged in the color palette for each index serving as the pair. By doing so, the index of the current pixel and the index of the adjacent pixel (before or after in the order of image processing) are close to each other, so that an improvement in the compression ratio can be expected.

Although the indexes are close in the pair of indexes determined in this way, the indexes are not always close between the pairs.
Therefore, assuming that the first color of interest is the color most frequently used in the image data, the index of the frequently used color and the color adjacent to the color are used. Because the pixel indices are close to each other,
An improvement in compression ratio can be expected. After determining each pair, the color used more frequently in the pair is set as the representative color, and the representative color is used most relatively in the image data. It is also conceivable to assign indices in order from the color. In this way, the index of the current pixel and the index of an adjacent pixel are close to each other, and the indexes of colors used more frequently are close to each other. Therefore, when viewed in the order of image processing, there is a high probability that values near the index continue, and an improvement in compression ratio can be expected.

In the case where the image data expresses an intermediate color by using a plurality of colors, the image data may be reconstructed so that the indexes of the plurality of colors forming the intermediate color are continuous. Since the indices of the colors used are close to each other, the continuity of the data is improved and the compression ratio can be expected to be improved despite the spatially random arrangement of the colors.

Incidentally, various types of image data are targeted. For example, in the case of a map image, claim 9
As shown in the figure, if the indices of the colors used for a specific element among the elements constituting the map image are close to each other, there are many characteristic elements in the map image. By assigning an index close to the element, data continuity is improved, and an improvement in compression ratio can be expected.

As the specific element, for example,
As shown in FIG. Although landmarks such as convenience stores and towers are displayed on the map image, the color scheme of these landmarks is often determined originally. By bringing the indices of colors used in landmarks close to each other, data continuity is improved,
An improvement in compression ratio can be expected.

Since the assignment of colors to the indices of the color palette is arbitrary, for example, if there are a plurality of indices corresponding to the same color, there is a possibility that an infinite number of indices as a whole cannot be used efficiently. Therefore, according to the eleventh aspect, the color palette can be used efficiently by sharing the colors used in the landmarks.

Further, the specific element may be a background and a road. Since the continuity of the map image in the background is often divided by roads, placing an index close to the background color index in the color used for the road improves the continuity of the data and reduces the compression ratio. Improvement can be expected. In this case, the road colors that divide the continuity of the background color can be efficiently arranged by sequentially arranging them in the index near the background color from the road color that is frequently used in the image data. , The continuity of the data is improved, and an improvement in the compression ratio can be expected.

Similarly, the specific element may be a background and a character. Since the continuity of the map image in the background is often divided by characters, placing an index close to the background color index in the color used for the characters improves the continuity of the data and reduces the compression ratio. Improvement can be expected. In such a case, by arranging the character colors frequently used in the image data in an index close to the background color sequentially, the road colors that divide the continuity of the background color can be efficiently set. , The continuity of the data is improved, and an improvement in the compression ratio can be expected.

Further, characters and roads can be cited as factors that divide the continuity of the background of the map image in this way. By arranging sequentially in the index near the background color from the color, road colors and character colors that divide the continuity of the background color can be efficiently arranged, improving the continuity of data and improving the compression ratio. Can be expected.

The functions of the data compression means, the coding means and the reconstructing means of the image compression apparatus to be realized by the computer system may be provided, for example, as a program started on the computer system side. Can be. In the case of such a program, for example,
Floppy (registered trademark) disk, magneto-optical disk, C
It can be used by recording it on a computer-readable recording medium such as a D-ROM or a hard disk, loading it into a computer system as needed, and starting up. Alternatively, the program may be recorded in a ROM or a backup RAM as a computer-readable recording medium, and the ROM or the backup RAM may be incorporated in a computer system and used.

[0023]

Embodiments of the present invention will be described below with reference to the drawings. It is needless to say that the embodiments of the present invention are not limited to the following examples, and can take various forms as long as they belong to the technical scope of the present invention.

FIG. 1 is a block diagram showing a schematic configuration of an image compression apparatus according to the embodiment. As shown in FIG. 1, the present image compression apparatus includes a reconstruction unit 10 and a data compression unit 20, and can perform compression processing on input multi-valued image data (an image before compression). Have been able to. Image data to be processed (compression target) by the image compression apparatus of the present embodiment is expressed in a color palette system, and compresses the index of the color palette. In the present embodiment, 256 colors can be allocated by using a value of 0 to 255 (decimal system) as the index. Therefore, the index has 8 bits when expressed in bits. Although the data compression unit 20 has the same configuration in the conventional device, the reconfiguration unit 10 is particularly characteristic.

First, the data compression section 20 will be briefly described. The data compression unit 20 includes a data conversion unit 21 and an encoding unit 22. The data conversion unit 21 decomposes the index into bit planes, performs binary run-length conversion on a plane-by-plane basis to increase the data bias, and sends the data to the encoding unit 22. The encoding unit 22 assigns a variable length code such as a Huffman code, for example. Note that, unlike the compression method for a normal direct color image, since lossless compression is essential, the amount of code is not controlled after encoding. RGB indicating color in direct color image
Even if irreversible compression is performed on the values, if the decoded RGB values are close to the original values, the colors are also close (FIG. 3).
(A)). However, since there is no correlation between the color palette index and the color information, if the index changes, even if the value is 1 (for example, 99 → 100), a completely different color (for example, 99 → 100) This is because there is a possibility that the color changes from red to blue (see FIG. 3B).

Therefore, prior to the processing in the data compression section 20, the reconstruction section 10 reconstructs the index of the color palette. That is, the color corresponding to the index is obtained by appropriately assigning the color included in the image data, and is intentionally reconstructed so as to be “suitable for data compression”. Specifically, the order of data processing in the image data (for example, the raster direction)
Are reconfigured so that the indices of consecutive pixels have similar values.

The meaning of such reconfiguration will be described. In the binary run-length conversion in the data conversion unit 21, since the number of consecutive data (that is, 0 or 1) is counted, the compression efficiency increases as the data continues. Therefore, how to make the data continuous is the point of improving the compression efficiency. When considering continuous pixels in image data, even if the color changes greatly (for example, red and blue), if the indexes of the colors that greatly change become close to each other, the bit There is a high possibility that the upper bits when the plane is decomposed become equal.

In the present embodiment, the index is represented by 8 bits. However, if the index is a continuous index, the index may be different only by the least significant bit. For example, when the continuous indexes 136 and 137 are considered, if they are expressed in bits, they are “10001000” and “10001”, respectively.
001 ”. In this case, only the least significant bit differs,
The upper 7 bits “1000100” are equal. Also,
When the continuous indexes 137 and 138 are considered, if they are represented by bits, "1000100"
1 "and" 10001010 ". Therefore, in this case, only the lower two bits differ, and the upper six bits “100
010 "will be equal.

Also, if the indexes are not consecutive but are numerically close, a significant number of upper bits will be equal in bit representation. Therefore, when binary run-length conversion processing is performed on the data of each plane obtained by decomposing the bit plane, the possibility that the same data continues is relatively high, and the encoding efficiency is relatively improved by encoding. .

It should be noted that if the indices are close to each other, it is true that "the possibility is high" that the upper bits become equal when the bit plane is decomposed. However, there is only a high possibility. For example, when each of the indexes is expressed in bits as in the case of 127 and 128, “01111111” and “1000000” are represented.
00 ", and there may be no bit that matches exactly.
However, since such a phenomenon occurs only locally, such as in the case of a carry that affects the upper bits,
Judging comprehensively, it can be said that there is a relatively high possibility that upper bits become equal when the bit plane is decomposed. However, in order to avoid such an inconvenience that there are few or no matching bits even if the index is close, the following measures may be taken. In other words, when reconstructing the indexes so that they are close to each other, (when the bit plane is decomposed)
The index is assigned such that the number of planes in which the data of the corresponding planes have the same value becomes relatively large. In other words, allocation is made to a portion such as 136 and 137 described above, which does not cause adverse effects due to carry. By doing so, the compression efficiency can be further improved.

As a basic idea, as described above, "reconstruction is performed so that the indices of consecutive pixels in the image data in the order of data processing are close to each other." If the value can be set as close as possible, the continuity of the data after the data conversion is improved, and further improvement in the compression ratio can be expected. Therefore, some ideas for making the indexes as close as possible will be described.

(1) The index of the color pallet colors being used is made continuous. Although there are 256 consecutive color palette indices from 0 to 255, it is conceivable that some of the indices are unused. In this case, if the indices that are not used are separated and the indices of the color palette colors that are used are made continuous, the indices of the above-described continuous pixels have a closer value. Also in this case, it is effective to add a device from the viewpoint of matching the values of the bit planes described above. For example,
As described above, the carry that affects the upper bits (for example, 01
111111 → 10000000) so that an index relationship such as “1111111 → 10000000)” does not occur. It should be added that similar measures can be taken in the following examples.

(2) Color pallet color indexes are sequentially assigned in correspondence with the order of colors having a large amount used in the image data. By doing so, the indexes of frequently used colors are close to each other, and the probability that consecutive index values are continuous in the image data is high, so that a further improvement in the compression ratio can be expected. For example, in the case of a map image, the amount of color used for the background is the largest,
Next, it is assumed that the color of the road is. In this case, since it is considered that the background and the road are often adjacent to each other, if the indices corresponding to these are made continuous, the continuity of the data after the data conversion is improved as described above. It should be noted that the color of the road often differs depending on the road type. For example, a plurality of road colors are set, and the amount of color used in the image is set in the background → a type A road (for example, a general road) → a type B road (for example, a national road) → a type C road → a type D road → Class E road ...
In this case, it is conceivable to assign indexes so as to be continuous in this order. By doing so, data that is likely to improve the continuity of the data after data conversion is more likely to occur in the order of the higher possibility, which contributes to the improvement of the compression efficiency as a whole.

(3) Assignment of Indexes When a large number of colors are used in an image, the following scheme can be considered for assigning which index to which color. Determining a Pair of Indexes A pair of two colors is determined by a method of determining the most color other than the color of interest as the index before or after the color of interest in the data before and after the color of interest. For example, considering the specific example of the above (2), first, a background color and a type A road color are paired. Next, a pair is determined recursively by a similar determination method for a color for which a pair has not been determined. That is, the color of the B type road and the color of the C type road are paired, and the color of the D type road and the color of the E type road are paired. Then, the colors are arranged in a color palette for each index of the colors to be paired. By doing so, the index of the current pixel and the index of the pixel before or after it (that is, the pixel adjacent to the current pixel in terms of the order of image processing) are close to each other, so that an improvement in compression ratio can be expected.

Consideration is also given to the pairs. The indexes are continuous within the pair of indexes determined as described above, but the indexes are not always continuous between the pairs. Therefore, after each pair is determined, the color that is used more frequently in the pair is set as the representative color, and an index is assigned in order from the color in which the representative color is used most frequently in the image data. For example, the background color and the type A road color are the first pair, the type B road color and the type C road color are the second pair,
Further, the color of the D type road and the color of the E type road form a third pair. Since the representative colors of these three pairs are the background color, the color of the B type road, and the color of the D type road, the order of the colors that are frequently used is the background color → the color of the B type road → It is the color of the D type road. Therefore, the first pair, the second pair, the third
Assign indices in pair order.

In this way, the index of the current pixel and the index of the adjacent pixel are close to each other, and the indexes of the colors used more frequently are close to each other.
Therefore, when viewed in the order of image processing, there is a high probability that values near the index continue, and an improvement in compression ratio can be expected.

(4) Handling of Dither Display Portion In the limited color expression using the color palette, in order to compensate for the lack of the number of colors, intermediate colors are created by so-called dither display at the expense of spatial resolution. For example, as shown in FIG. 4, intermediate colors are used in the background, parks, forests, and the like. Therefore, the index between adjacent colors is set to a value close to this dither display.

If there is no correlation between indices corresponding to colors used for dither display, data becomes random even if bit plane decomposition is performed as shown in FIG. But,
As shown in FIG. 5B, when the index is a close value, it can be seen that the higher plane has less fluctuation and the continuity is improved. In the case of dither display, since the frequency of color change is particularly large, if no countermeasures are taken, there is a high possibility that the discontinuity of data after bit plane decomposition is increased. Therefore, when considering the improvement of the compression ratio of the entire image, it is important to deal with this part.

(5) Dealing with Landmark Display Portion In this embodiment, a map image is to be compressed, and may be an image corresponding to a landmark display such as a convenience store or a tower (FIG. 6A). reference). Therefore, the indexes of the colors used in this landmark are close to each other. Since the color scheme of landmarks is often determined in the first place, if the color palette is reconfigured so that the indices of the colors used for landmarks are close to each other, the continuity of data is improved, and the compression ratio is improved. Can be expected to improve.

Since the assignment of the colors to the indexes of the color palette is arbitrary, for example, if there are a plurality of indexes corresponding to the same color, there is a possibility that the limited number of indexes cannot be used efficiently as a whole. is there. For example, there are 256 indices, but if the same color is assigned to each of the two indices, only half of the 128 colors can be used. Therefore, if the colors used for the landmarks are shared, the color palette can be used efficiently. (See FIG. 6B).

(6) Dealing with Other Parts The landmark described in the above (5) is an example of a specific element used in the map image. In other words, the idea is that the indices of the colors used in the specific element are close to each other.
Since landmarks are generally composed of a plurality of colors, the indices of colors used in the landmark are close to each other. However, specific elements may be, for example, a background and a road, a background, and characters. . This is because the continuity of the map image in the background is often divided by roads and characters.

In this case, a plurality of colors are often used as the color of the road and the color of the character. Therefore, in that case, the colors may be arranged in the index near the background color sequentially from the color of the road and the color of the character which are frequently used in the map image.
For example, in the case shown in FIG. 6C, since the usage is large in the order of general road → national road → character → toll road, the index of the color of the general road and the color of the national road are arranged next to the index of the background color, respectively. In addition, the character and color index of the toll road are arranged next to them. In this way, it is possible to efficiently arrange road colors and character colors that divide the continuity of the background color, thereby improving the continuity of data and improving the compression ratio.

[Others] (1) The present applicant has proposed the following image compression technique in Japanese Patent Application No. 2000-130177. For example, taking a map image as an example, any color in the image (for example, background)
To separate it into a complementary color file and a selected color flag file, and compress them separately. The complementary color file is data in which the background portion is the same as the color of the road, and the background color is a color that is used relatively frequently in the map image data. Then, in the image whose background is the same color as the road, the continuity of the same color is considerably increased with respect to the original image. In other words, the continuity of the data was interrupted because the color changed at the switching portion from the background to the road or vice versa, but the continuity was increased by doing so. for that reason,
When run-length conversion is performed, the continuous length of the same data becomes relatively long, and the compression efficiency of such an unseparable original image is improved. As a result, even if the selected color flag file is separately compressed, the compression efficiency is improved as compared with the case where the original file is directly compressed. This technique can be combined with the present invention. The creation of the complementary color file in this case may be performed before or after the reconstruction of the color palette, but is preferably performed before the reconstruction.

Note that the above-mentioned Japanese Patent Application No. 2000-130177 is disclosed.
Discloses that a file after thinning may be used instead of the above-described complementary color file. In other words, in the case of the complementary color file, the color is complemented with the color of the road, for example, instead of the selected background color, but in the case of the file after thinning, no complement is performed simply by thinning out (extracting) the background color. . A file for indicating the thinned position is required, but such a method can be similarly combined.

(2) Although a map image is taken as an example of image data to be compressed, the present invention is not limited to this, and may be, for example, a natural image. However, in the case of a map image, the same color continues in the background area, and when the color changes, it is steep and the high frequency component of the spatial frequency is large. Therefore,
By improving the continuity of the upper bits when the bit plane is decomposed as in the present invention in the color change portion, it is easy to expect an improvement in the compression efficiency of the entire image.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an image compression apparatus according to an embodiment.

FIG. 2 is an explanatory diagram of a direct color system and a color pallet system.

FIG. 3 is an explanatory diagram of a difference between lossy compression of a direct color method and color pallet method.

FIG. 4 is an explanatory diagram of dither display in a map image.

FIG. 5 is an explanatory diagram of a cause of a reduction in the compression ratio and a concept for improving the compression ratio in a color palette system.

6A is an explanatory diagram of a map image on which landmarks are displayed, FIG. 6B is an explanatory diagram of an efficient use method of a color palette, and FIG. 6C is an explanatory diagram of an element in the map image. FIG. 9 is an explanatory diagram of a color palette reconfiguration example for colors.

[Explanation of symbols]

10 Reconstruction unit, 20 Data compression unit, 21 Data conversion unit, 22 Encoding unit

Claims (18)

[Claims] 1. Data conversion means for decomposing multi-valued image data represented by a color palette system into bit planes and sequentially or serially performing binary run-length conversion processing on the data of each plane; An image compression apparatus comprising: an encoding unit that assigns a code to data converted by the conversion unit. Prior to the conversion by the data conversion unit, indexes of consecutive pixels in the order of data processing in the image data are close to each other. An image compression apparatus comprising a reconstructing unit for reconstructing a color palette so as to obtain a value. 2. The image compression apparatus according to claim 1, wherein the reconstructing unit is configured to determine the number of planes in which data of corresponding planes has the same value with respect to the index to be reconstructed to have the near value. An image compression apparatus, wherein an index is assigned such that the number of images is relatively large. 3. The image compression apparatus according to claim 1, wherein said reconstructing means is configured such that indices of color pallet colors used in said image data are continuous and continuous in said image data. An image compression apparatus reconstructing a color palette so that pixel indices are close to each other. 4. The image compression apparatus according to claim 1, wherein said reconstructing means includes a color palette color corresponding to an order of a color having a large amount used in said image data. An image compression apparatus characterized by sequentially assigning indices. 5. The image compression apparatus according to claim 1, wherein the reconstructing unit determines a most-recent color other than the focused color in data before and after the focused color. A pair of two colors is determined by a method of determining the index before or next to the color, and for a color for which no pair is determined, a pair is determined recursively by the above-described determination method, and a color palette is determined for each pair. An image compression apparatus, wherein: 6. The image compression apparatus according to claim 5, wherein the reconstructing unit sets the first color of interest to a color that is used most frequently in the image data. Image compression device. 7. The image compression apparatus according to claim 6, wherein after determining each pair, the reconstructing means sets a color used more frequently in the pair as a representative color, and sets the representative color as the representative color. An image compression apparatus, wherein an index is assigned in order from a color that is used most frequently in image data. 8. The image compression apparatus according to claim 1, wherein said reconstructing means creates an intermediate color when said image data expresses an intermediate color using a plurality of colors. An image compression apparatus for reconstructing an index of a plurality of colors so as to be continuous. 9. The image compression apparatus according to claim 1, wherein said reconstructing means comprises: when the image data is a map image,
An image compression apparatus characterized in that, among the elements constituting the map image, the indices of the colors used for the specific elements have similar values. 10. The image compression apparatus according to claim 9, wherein said specific element is a landmark displayed on a map image. 11. The image compression apparatus according to claim 10, wherein, when a common color exists between two landmarks, the reconstructing unit uses a color other than the common color used in one of the landmarks. Are arranged so as to be continuous with the index before the common color, and the colors other than the common color used in the other landmark are arranged so as to be continuous with the index after the common color. Characteristic image compression device. 12. The image compression apparatus according to claim 9, wherein said specific elements are a background and a road. 13. The image compression apparatus according to claim 12, wherein said reconstructing means sequentially arranges the colors of the roads used in said image data in an index near a background color. Image compression device. 14. The image compression apparatus according to claim 9, wherein said specific elements are a background and a character. 15. The image compression apparatus according to claim 14, wherein said reconstructing means sequentially arranges in an index near a background color from a character color frequently used in said map image. Image compression device. 16. The image compression apparatus according to claim 9, wherein said reconstructing means arranges in an index near a background color sequentially from a character or road color frequently used in said map image. An image compression device characterized by the above-mentioned. 17. The image compression apparatus according to claim 1, wherein said encoding means assigns a variable length code to the data converted by said data conversion means. 18. A computer-readable recording medium in which a program for causing a computer system to function as a data conversion means, an encoding means, and a reconstruction means of the image compression apparatus according to any one of claims 1 to 17 is recorded.