JP2005134601A - Analysis system for map image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To extract the differences between a plurality of map image data with good accuracy. <P>SOLUTION: The maps of a new version and an old version which are objects for comparison are read with a scanner 10 and respective pieces of the image data are obtained in the form of raster data of multicolors. Next, the points of triangulation, the angles of diagram frames and the positions of registry guides are extracted as the corresponding points existing in both of the image data. Further, the maps are divided to Delaunay triangles based on the corresponding points and are subjected to compensation by each of the regions in such a manner that the coordinates of the vertexes of the respective triangular regions coincide with each other. Pieces of the image data subjected to the compensation in such a manner are compared with each other in pixel units and the differences between the new and old maps are extracted. In these process steps, the amount of information is reduced by converting the color information included in the maps to gray scales and the errors included in the color information are eliminated by quantization. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an image analysis for extracting a difference between a plurality of maps for a map image.

  Since the buildings and terrain are constantly changing, the map will be modified accordingly, and new versions of the map will be published as appropriate. In recent years, since digitization of maps has progressed, it is necessary to reflect such changes in digital data. However, new maps are not always available in digital data. In order to facilitate digital data correction when a new version of a map is obtained in the form of a printed matter, a technique for automatically extracting a difference from an old version of the map has been proposed. For example, Patent Document 1 proposes a device that reads new and old maps with a scanner and converts them into digital data, and automatically extracts the difference in shape between the two. In this apparatus, in order to improve the accuracy of difference extraction, it is possible to align the scan images of the old and new maps based on the operation of the operator.

JP-A-9-319855

  Various problems still remain in extracting the difference between the old and new maps. For example, with the technique of Patent Document 1, the work burden on the operator for performing alignment is large. In addition, the accuracy of alignment depends on the operation of the operator, and there is a problem that sufficient accuracy is not always guaranteed. In the conventional technology, differences in the part where various objects such as letters and contour lines are printed are overlapped, and differences that do not appear clearly in the shape of the object, such as when a building with almost the same shape as the land is formed in a vacant place The extraction accuracy was not always sufficient. Such a problem is common when a difference between a plurality of maps is extracted regardless of old and new maps. The present invention has been made in view of such a problem, and an object of the present invention is to reduce the workload and improve the accuracy when extracting differences between a plurality of maps.

  The present invention can be configured as an analysis device that extracts a difference between two maps. It is also possible to extract the difference between three or more maps by repeatedly applying the difference extraction for the two maps.

  The analysis apparatus of the present invention inputs image data representing two maps from which differences are to be extracted. The analysis apparatus itself may be provided with a scanner for reading a map as a printed matter, or may be configured to input separately digitized data. Based on the image data input in this way, the analysis apparatus extracts a predetermined corresponding point that exists in common in the two maps. Further, the image data is divided into a plurality of areas defined by corresponding points, and the image data is corrected for each area so that the positions of the corresponding points coincide. Further, the difference between the two maps is extracted based on the corrected image data.

  According to the analysis apparatus of the present invention, the corresponding points are automatically extracted without being designated by the operator, and the image data is corrected. Can be performed accurately.

  As the corresponding points, various characters, symbols, figures, etc. printed on the map (hereinafter collectively referred to as “objects”) can be used. For example, a reference point used for surveying at the time of map creation, such as a triangle point or a level point, may be used as a point that is surely present on the map and whose exact latitude and longitude can be easily specified. In addition, as long as it is indicated by a dot, a building such as a station, a landmark, or an administrative boundary may be used. In addition to the object, the corresponding points can be obtained by using corners of the map frame of the map, internal dividing points of a specific ratio such as the center points on the sides of the map frame, and reference lines outside the map frame called register marks. Good.

  The area serving as the correction unit of the image data can be arbitrarily set in various shapes based on the corresponding points. Corresponding points need not necessarily be the vertices of the region. However, in order to ensure the accuracy of correction in the variously settable areas, a triangle having a corresponding point as a vertex is preferable, and a Delaunay triangle is preferably applied. A Delaunay triangle refers to a shape that is set to satisfy the condition that the vertex of another triangle does not enter the circumscribed circle of any triangle.

  As the image data, raster data having a gradation value for each pixel can be used. In this case, when the image data is corrected, a pixel whose gradation value is not uniquely determined may appear. For example, there are a case where a plurality of pixels are originally associated with one pixel by correction, and a case where a pixel which is not associated with any pixel before correction appears. For such a pixel, the gradation value may be obtained by the following procedure. First, the position corresponding to each pixel after correction is obtained in the image data before correction. Next, a gradation value at the position is obtained based on a predetermined calculation using a plurality of pixels existing in the periphery of the image data before correction. As the calculation, for example, linear interpolation or a method of employing the nearest gradation value can be employed. The gradation value thus obtained is set as the gradation value of the corrected pixel. By doing so, it is possible to reduce noise generated based on the correspondence between the pixels before and after correction, and improve the accuracy of extracting the difference.

  Multi-color data may be used as the image data. In such a case, it is preferable to extract the corresponding points based on both the shape and color of the corresponding points. By doing so, the extraction accuracy of corresponding points can be improved. For example, even when an object other than the corresponding point such as a contour line or a character is printed on the corresponding point, the corresponding point can be accurately extracted as long as both colors are different. In addition, it is possible to reduce the possibility that an object whose shape matches the shape of the corresponding point but has a different color is mistaken as a corresponding point.

  When using multi-color image data, the analysis apparatus of the present invention may extract a portion where at least one of a change in shape and a change in color occurs for each object as a difference between two maps. Such an extraction method can be applied regardless of whether or not the above-described image data correction is performed. According to this extraction method, for example, a difference that does not appear clearly in the shape of the object, such as when a building having almost the same shape as the land is formed in a vacant place, is accurately extracted based on the change in the color of the object. It becomes possible to do.

  When multicolor image data is used, extraction may be performed after the image data is separated based on the color information of the image data. For example, when contour lines are printed in brown, image data from which contour lines have been deleted can be obtained by separating the image data into image data of brown and other colors, and difference extraction is performed for objects that overlap the contour lines. It becomes easy. As described above, by using the color separation process based on the color information, it is possible to improve the accuracy of extracting the difference between the objects printed in an overlapping manner.

  The color information is usually represented by a plurality of components such as the three primary colors of red, green, and blue. In the present invention, it is preferable to perform color correction that reduces the number of these components based on a predetermined calculation. . Such color correction includes, for example, correction of so-called gray scale from data of three primary colors. Since the number of colors used for the map is usually relatively small, even if the number of components is reduced in this way, the color difference for each object is maintained. By reducing the number of components based on such map characteristics, it is possible to reduce the amount of data for representing color information while ensuring color distinction.

  The color information may be further quantized based on the reduced component value. The quantization can be performed by, for example, a method of assigning the same component value to colors whose component values are within a certain range. This corresponds to a correction for considering the same color when there is a slight difference in color information. As described above, since the number of colors used in the map is relatively small, it is possible to ensure the usefulness of the color information, that is, to distinguish colors for each object, even if such quantization is performed. Quantization can suppress the influence of noise generated in color information during printing or digitization while ensuring the usefulness of color information in this way, and improve the accuracy of difference extraction. There are advantages.

  In the analysis apparatus of the present invention, all of the various features described above do not necessarily have to be applied, and can be applied by appropriately combining or selecting them. The present invention may be configured as an analysis method for extracting a map difference by a computer in addition to the configuration as an analysis apparatus. Further, the present invention can be configured as a computer program for causing a computer to perform such analysis, or a computer-readable recording medium on which such a computer program is recorded.

  As a recording medium, a flexible disk, a CD-ROM, a magneto-optical disk, an IC card, a ROM cartridge, a punch card, a printed matter on which a code such as a barcode is printed, an internal storage device of a computer (RAM or ROM) Etc.) and various media that can be read by a computer, such as an external storage device.

Embodiments of the present invention will be described in the following order.
A. Device configuration:
B. Map analysis processing:
B1. Preprocessing routine:
B2. Corresponding point extraction processing:
B3. Alignment process:
B4. Difference extraction processing:

A. Device configuration:
FIG. 1 is an explanatory diagram showing a schematic configuration of a map analyzing apparatus as an embodiment. The analysis apparatus has a hardware configuration in which a personal computer 100, a scanner 10, and a hard disk 20 are connected. This analyzer automatically extracts differences between maps, that is, differences in topography and features based on analysis of image data obtained by reading the map MAP2 obtained in the form of printed matter with the scanner 10. It is. For example, if a new version of a map is obtained in the form of a printout when digitized map data is obtained for an old version of the map, the difference between the two is extracted to reduce the workload required to update the map data. Can be used to The hard disk 20 is used to store difference extraction results and map data. An old version of the map may be used to store scanned image data. The analysis apparatus can be generally used to extract a difference between a plurality of maps, but for the sake of convenience of explanation, an explanation will be given below using a comparison between old and new maps.

  The function as an analysis apparatus is realized by installing a predetermined program in the personal computer 100. In the figure, functional blocks configured by software by this program are shown. At least a part of these functional blocks may be configured as hardware, or may be provided by distributed processing in cooperation with another computer or server.

  The image data input unit 110 inputs image data to be analyzed. Both the old and new maps may be read by the scanner 10, or at least one of the data may be input from the hard disk 20 or the like. When digitized data is prepared for both old and new, the analysis apparatus may have a hardware configuration in which the scanner 10 is omitted. Although the format of the image data can be arbitrarily selected, in this embodiment, the image data includes 8-bit gradation values for each pixel for the three primary colors of red (R), green (G), and blue (B). The description will be made assuming that the defined raster data is used. It is assumed that the map of this embodiment is printed in black, blue, brown and multicolor. In this case, the map image is an image including four colors including the white background. However, each color does not always have the same gradation value due to color unevenness or other causes at the time of printing or digitization, and actual image data has more color data than the above four colors. .

  The input image data contains misalignments of the printed material itself and distortions that occur during the digitization process. For accurate comparison, corrections such as alignment and distortion correction are required. Preferably it is done. The corresponding point extraction unit 120 extracts corresponding points from each image data as a reference used for the correction. The position of the corresponding point is specified by a coordinate system set on the image data. An image to be extracted as a corresponding point is defined in advance by the template 121. In this embodiment, a triangular point, a corner of a map frame, and a reference line called a register mark are used as corresponding points.

  On the map, contour lines, various features, and the like may be printed over corresponding points. In the present embodiment, a separation process based on color information is applied in order to accurately extract corresponding points for such locations. In the present embodiment, the color separation processing refers to processing for removing or extracting any specific color portion from map image data. The separation processing unit 160 executes this separation processing.

  The color correction unit 170 may be referred to as color correction processing (hereinafter referred to as “grayscale processing”) for reducing color information of image data from three RGB components to a single component in order to facilitate color separation processing. ) And an error caused by color unevenness is eliminated, and a quantization process for classifying the color into one of four colors is executed. The correction value table 171 gives correction formulas used for color correction processing, reference values used for quantization processing, and the like.

  The alignment unit 130 corrects the image data of the old and new maps so that the coordinates of the corresponding points extracted by the corresponding point extraction unit 120 match. The corrected image data is managed by the correction data management unit 131. The difference extraction unit 140 compares the new and old image data after correction, and extracts the difference between the two. Also at the time of this extraction, a separation process is applied as appropriate to improve accuracy. The output unit 150 stores the difference thus obtained in the hard disk 20 as difference data.

  FIG. 2 is an explanatory diagram showing a map example to be analyzed. From the top, the new map, the old map, and the differences between them are shown schematically. The land is represented by a white background. It is assumed that the sea located at the upper left (dashed hatched portion) is printed in blue. Contour lines shall be represented in brown. In the figure, contour lines are represented by alternate long and short dash lines in order to express color differences. Other roads, buildings (hatched areas), letters, triangles, etc. are printed in black.

  In this example, old and new differences are generated in areas A to E shown in the new version map. In the difference map, an output format is shown in which a difference is generated on a map in which the old and new versions are overlapped and expressed in a color different from the four colors used for printing the map, such as red. For convenience of illustration, the difference portions are shown by cross hatching in the drawing. The difference map does not need to be superimposed on the old version or new version of the map, and only the difference portion may be output.

B. Map analysis processing:
FIG. 3 is a flowchart of the map analysis process. It is a process for extracting the difference between the old and new maps, and is a process executed by the CPU of the personal computer 100 constituting the analysis apparatus. When the process is started, the CPU inputs image data of old and new maps (step S100), and performs color correction, quantization, and noise removal as preprocessing for extracting the difference (step S200).

  Next, the CPU executes the corresponding point extraction process described above on the image data obtained in this way (step S300), and performs the alignment process based on the corresponding point, that is, the old and new so as to match the coordinates. A process of correcting the image data of the map image data is performed (step S400). Based on the correction data thus obtained, the CPU extracts the difference between the two (step S500) and outputs the result (step S600). Hereinafter, each process will be described in detail.

B1. Preprocessing routine:
FIG. 4 is a flowchart of the preprocessing routine. This is a process corresponding to step S200 of the analysis process (FIG. 3).

In this process, the CPU performs color correction and quantization for each pixel of the image data (step S210). As color correction, the CPU first converts the RGB gradation values of each pixel into two component values GD1 and GD2 according to the following equation.
GD1 = 0.596 * R−0.274 * G−0.322 * B;
GD2 = 0.149 * R + 0.437 * G + 0.414 * B
GD1 is a value corresponding to a color difference signal from orange to cyan in the YIQ color system. GD2 is a parameter representing the brightness difference of the RGB color system.

Based on the data thus obtained, the CPU classifies each color of the image data into blue (B), brown (Br), white (W), and black (K). In this embodiment, an 8-bit gray scale GS is given as an index value representing each color. The classification conditions and gray scale values for each color were set as follows.
−55 <GD1 <−11 → GS = 200 (blue: B);
9 <GD1 <50 → GS = 120 (Brown);
200 <GD2 → GS = 250 (white: W);
GD2 <140 → GS = 0 (black: K)

  In addition, the state of classification of each color is shown in the figure. The parameter GD1 is used for identifying blue and brown, and GD2 is used for identifying white and black. That is, when the conversion formula shown above is used, the blue pixels in the image data are distributed in the range of “−55 <GD1 <−11” for GD1, although the parameter GD2 takes various values. To do. Therefore, these pixels are quantized, that is, the difference between the values in the parameters GD1 and GD2 is ignored, and they are treated as the same color, and 200 representing blue (B) is used as the GS value. give.

  Similarly, for a pixel whose parameter GD1 is “9 <GD1 <50”, 120 representing brown (Br) is uniformly given as the GS value. White and black are evaluated based on the parameter GD2, and the above-described GS values are given, that is, 250 for white and 0 for black, respectively.

  By this processing, the color information of the image data can be expressed by a single parameter called gray scale GS, and various errors included in the color information can be eliminated. By doing so, it is possible to facilitate the handling of color information by taking advantage of the fact that the map is printed with several known colors in advance.

  The correction formula or parameter used for color correction and quantization, and the reference value for quantization are preferably used according to the color used in the map. In this embodiment, these correction equations are given by the correction value table 171 described above with reference to FIG. In this way, for example, by specifying the number of colors used in the map prior to analysis, the operator can use different correction formulas and the like, and appropriate color correction and quantization can be performed. It becomes possible. In the embodiment in the figure, two types of parameter GD1 for identifying blue and brown and parameter GD2 for identifying white and black are used. However, evaluation is performed using a single parameter or three or more parameters. It doesn't matter.

  The CPU performs noise removal on the data thus subjected to color correction and quantization (step S220). Although various methods can be applied to the noise removal, in this embodiment, the noise removal is performed by two types of methods: median filter and diffusion / shrinkage processing.

  The median filter processing method is schematically shown in the figure. The squares in the figure indicate pixels, and it is assumed that hatched pixels and white pixels have different colors. When the processing target pixel for noise removal is P1, the median filter pays attention to nine pixels including the periphery (portion surrounded by a thick line in the figure), and the median value of the gradation values of these pixels is the target pixel P1. Gradation value. In the example shown in the figure, when nine pixels are arranged in order of gradation values, the median value, that is, the fifth pixel is a hatched pixel, and therefore this gradation value is assigned to the target pixel P1.

  In the figure, the diffusion / shrinkage treatment method is schematically shown. The contraction process refers to a process of replacing a target pixel with a blank pixel when there is a blank pixel in at least a part of the surrounding eight pixels of the target pixel. According to this processing, the scattered noise is eliminated as shown at the left end in the figure.

  Contrary to the diffusion process, when there is a non-blank pixel having a gradation value in at least a part of the surrounding eight pixels of the blank target pixel, the process of replacing the target pixel with the non-blank pixel is performed. say. According to this processing, noise having blank pixels in a dot shape as shown in the lower right in the figure is eliminated.

B2. Corresponding point extraction processing:
FIG. 5 is a flowchart of corresponding point extraction processing. This is a process corresponding to step S300 of the analysis process (FIG. 3). When the process is started, the CPU reads a template of corresponding points (step S310). A part of the map is shown in the figure, and an example of the template is shown. The hatched part in the figure corresponds to the inside of the map frame.

  As described above, in this embodiment, the triangular point, the corner of the picture frame, and the registration mark are used as corresponding points. As the template T1 representing a triangular point, image data of a certain region including the triangular point is used. For the template T1, the portions other than the triangular points may be blank or white. Image data including these parts is used as a template T2 representing a corner of a picture frame and a template T3 representing a registration mark. As for the templates T2 and T3, any gradation value may be used for the image inside the frame and the hatched portion.

  Next, the CPU extracts corresponding points by template matching (step S320). At this time, in order to improve the extraction accuracy of corresponding points, a color separation process is performed. The state of separation processing and template matching is illustrated in the figure. In the example shown in the figure, the contour line image PL2 and the other image PL1 are separated by performing color separation processing with brown and other colors. In this way, the degree of coincidence between each pixel constituting the template T1 and the image PL1 is obtained at each position while sequentially moving the template T1 on the image PL1. The degree of coincidence may be evaluated only for the pixel portion forming the triangular point in the template. It can be determined that a triangular point as a corresponding point exists in a place where the degree of coincidence obtained in this way is a predetermined value or more.

  The same applies to other templates. While moving the templates T2 and T3, it is determined that corners and registration marks exist where the degree of coincidence increases. Regarding the templates T2 and T3, portions corresponding to images in the figure frame may be evaluated as matching with the template if the image PL1 is other than 0.

  The CPU specifies the coordinates of the corresponding point for each of the image data of the old and new maps through the above processing. In this embodiment, x and y coordinates with the origin at the lower left of the map are used.

B3. Alignment process:
FIG. 6 is a flowchart of the alignment process. This is a process corresponding to step S400 of the analysis process (FIG. 3).

  The CPU first divides the image data into Delaunay triangles (step S410). A Delaunay triangle refers to a shape that is set to satisfy the condition that the vertex of another triangle does not enter the circumscribed circle of any triangle. The state of division is shown in the figure. Triangular points are shown as filled triangles. The hatched area is one of the Delaunay triangles. The area division of the image data is not limited to the Delaunay triangle, and any shape can be applied. However, the use of Delaunay triangles has the advantage that correction of image data, which will be described later, can be realized relatively easily and accurately.

  The CPU performs correction for each region so that the vertices of the triangle thus obtained match the new and old image data. For this correction, various methods can be applied. In this embodiment, affine transformation is performed. In order to perform this conversion, the CPU determines an affine transformation coefficient (step S412). The image of affine transformation is shown in the figure. Assume that the vertices of the Delaunay triangle in the old version of the map are A, B, and C, and the vertices a, b, and c of the Delaunay triangle in the new version of the map. The affine transformation is a transformation realized by a combination of parallel movement, enlargement / reduction, and rotation, and has a feature that the x and y coordinates before and after the transformation are expressed by a linear expression. The CPU can set these coefficients so that the vertices a, b, and c of the new version of the Delaunay triangle coincide with the vertices A, B, and C of the old version. In this embodiment, the image data of the new version map is corrected. However, the image data of the old version map may be corrected, or both may be corrected.

  When the coefficient of affine transformation is set in this way, the CPU first selects one unprocessed pixel included in the image data of the Delaunay triangle after the conversion as a processing target pixel (step S414). Then, by performing inverse affine transformation on the target pixel, a coordinate value in the image data before conversion is calculated (step S416). Next, the GS value in each pixel of the image data before conversion is interpolated to obtain a GS value corresponding to this coordinate value (step S418).

  The state of interpolation calculation is shown in the figure. Each square in the figure represents a pixel in the image data before conversion, and the coordinates obtained in step S416 are represented by a point P. In the example shown in the figure, around the point P, there are four pixels shown in white. Therefore, the GS value of the point P can be obtained by interpolating the GS values of these four pixels, GSa to GSd. Since such two-dimensional data interpolation is a well-known calculation, a detailed description thereof will be omitted. The CPU assigns the GS value thus obtained to the target pixel. The CPU repeats the above processing for all the target pixels after conversion (step S420).

  The above correction method has the following advantages. Generally, when image data is subjected to affine transformation, pixels before and after the transformation do not always correspond one-on-one. A plurality of pixels before conversion may be associated with one pixel after conversion, or none of the pixels may be associated with each other. In such a case, the GS value of the pixel after conversion is not uniquely determined. On the other hand, according to the method described above, by applying the interpolation calculation, the GS value corresponding to each pixel after conversion can be uniquely obtained, and there is a possibility that an error or noise of the GS value accompanying the conversion occurs. Can be suppressed.

  The conversion method shown in the embodiment is merely an example, and various other methods can be applied to the conversion of image data. In step S418, quantization of the GS value may be performed after the interpolation calculation, or color correction and quantization may be performed after the interpolation calculation is performed with RGB gradation values.

B4. Difference extraction processing:
FIG. 7 is a flowchart of the difference extraction process. This is a process corresponding to step S500 of the analysis process (FIG. 3). In the present embodiment, for each pixel of the image data of the new version map, the difference is extracted by comparing whether or not it matches the image data of the old version map.

  In order to execute such processing, the CPU first selects one of unprocessed pixels from the image data of the new version map as a processing target pixel (step S502). Next, the new and old image data are compared for this pixel (step S504). The comparison is schematically shown in the figure. The squares in the figure represent pixels. The image data of the new version map is shown on the upper side, and the image data of the old version map is shown on the lower side. Filled cells are pixels to be processed. In this embodiment, the target pixels are not only the pixels at the corresponding positions in the image data of the old version map (filled pixels in the figure) but also the eight neighboring pixels (hatched pixels in the figure). To be compared.

  When the GS value matches with at least one of these pixels (step S506), it is determined that there is no difference between the old and new for the target pixel, and the value 0 is stored as data representing the difference (step S508). On the other hand, if the GS value does not coincide with any of the above-described pixels (step S508), it is determined that there is an old / new difference for the target pixel, and a value 1 is stored as data representing the difference (step S510). The CPU repeats the above processing until all the pixels are finished (step S512).

  In this comparison, the color difference may be taken into account or ignored. When considering the difference in color, for example, when there is a building with almost the same shape as the land in a vacant place, or when the pond is reclaimed into land, there will be no change in shape. There is an advantage that the difference can be extracted as a difference based on the color information.

  In this embodiment, as described above, by comparing not only the target pixel but also its surrounding pixels, the difference is accurately recognized even when a difference of several pixels remains in the image data of the old and new maps. can do. The narrower the region to be compared, the higher the sensitivity of difference extraction, and the wider the region, the lower the sensitivity. The area to be compared can be arbitrarily set according to the required sensitivity for difference extraction and the accuracy of image data and alignment. For example, peripheral pixels may not be included in the comparison target.

  According to the analysis apparatus of the present embodiment described above, the corresponding points can be automatically extracted, so that the new and old maps can be accurately aligned with a light load. In addition, by utilizing color information, it is possible to improve the accuracy of extraction of corresponding points and difference extraction. Furthermore, in these processes, the color information is reduced to a single component called a GS value, and by handling with quantization, errors contained in the color information can be eliminated, and the color information can be effectively used with a light load. It becomes possible to do.

  As mentioned above, although the various Example of this invention was described, it cannot be overemphasized that this invention is not limited to these Examples, and can take a various structure in the range which does not deviate from the meaning.

It is explanatory drawing which shows schematic structure of the map analysis apparatus as an Example. It is explanatory drawing which shows the example of a map used as analysis object. It is a flowchart of a map analysis process. It is a flowchart of a pre-processing routine. It is a flowchart of a corresponding point extraction process. It is a flowchart of an alignment process. It is a flowchart of a difference extraction process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Scanner 20 ... Hard disk 100 ... Personal computer 110 ... Image data input part 120 ... Corresponding point extraction part 121 ... Template 130 ... Positioning part 131 ... Correction data Management unit 140 ... Difference extraction unit 150 ... Output unit 160 ... Separation processing unit 170 ... Color correction unit 171 ... Correction value table

Claims (12)

An analysis device that extracts the difference between two maps,
An image data input unit for inputting image data representing the two maps;
Based on the image data, a corresponding point extracting unit that extracts a predetermined corresponding point that exists in common in the two maps;
An image correction unit that divides the image data into a plurality of regions defined by the corresponding points and corrects the image data for each of the regions so that the positions of the corresponding points match;
An analysis apparatus comprising: a difference extraction unit that extracts a difference between the two maps based on the corrected image data. The analysis device according to claim 1,
The corresponding point includes an analysis device including at least a part of a reference point used for surveying at the time of map creation, a corner of a map picture frame, an internal dividing point of a specific ratio on each side of the map picture frame, and a registration mark. The analysis device according to claim 1,
The image data is raster data having a gradation value for each pixel,
The image correction unit obtains a position corresponding to each pixel after correction in the image data before correction, and based on a predetermined calculation using a plurality of pixels existing around the position in the image data before correction. An analysis apparatus that obtains a gradation value at the position and sets the gradation value as a gradation value of the pixel after correction. The analysis device according to claim 1,
The image data is multicolor data,
The corresponding point extraction unit is an analysis device that extracts the corresponding points based on both the shape and color of the corresponding points. An analysis device that extracts the difference between two maps,
An image data input unit for inputting multicolor image data representing the two maps;
For each object represented in the image data corresponding to the two maps, a difference extraction unit that extracts a portion where at least one of a change in shape and a change in color occurs as a difference between the two maps; Analyzing device provided. The analysis device according to claim 5,
The differential extraction unit is an analysis device that performs the extraction after separating the image data based on color information of the image data. The analysis device according to any one of claims 4 to 6,
The image data is multicolor image data,
An analysis apparatus including a color correction unit that reduces the number of components used to represent the color based on a predetermined calculation. The analysis device according to claim 7,
The color correction unit is an analysis apparatus that quantizes color information included in the image data based on the reduced component value. An analysis method for extracting the difference between two maps,
As a process executed by the computer,
Inputting image data representing the two maps;
A step of extracting a predetermined corresponding point existing in common in the two maps based on the image data;
Dividing the image data into a plurality of regions defined by the corresponding points, and correcting the image data for each region so that the positions of the corresponding points coincide;
And a step of extracting a difference between the two maps based on the corrected image data. An analysis method for extracting the difference between two maps,
As a process executed by the computer,
Inputting multicolor image data representing the two maps;
Analyzing each object represented in the image data corresponding to the two maps by extracting a portion where at least one of a change in shape and a change in color occurs as a difference between the two maps Method. A computer program for extracting the difference between two maps,
A function of inputting image data representing the two maps;
Based on the image data, a function of extracting a predetermined corresponding point that exists in common in the two maps;
A function of dividing the image data into a plurality of regions defined by the corresponding points and correcting the image data for each of the regions so that the positions of the corresponding points coincide;
A computer program for realizing by a computer a function of extracting a difference between the two maps based on the corrected image data. A computer program for extracting the difference between two maps,
A function of inputting multicolor image data representing the two maps;
A function of extracting at least one of a change in shape and a change in color of each object represented in the image data corresponding to the two maps as a difference between the two maps by a computer Computer program to realize.

Images