JPH07287751A - Method for recognizing house graphic from map segment data - Google Patents

Abstract

PURPOSE:To improve probability of recognizing a house part from map segment data by searching a segment pair forming a rectangular corner from among segment data, listing up visual segments as start points of recognition processing, finding out the number of corners almost perpendicular to an interval between the end points of a segment sequence, and recognizing a house graphic based upon the found number of corners. CONSTITUTION:Segment data obtained by reading out a town map data prepared by a scale of less than a 1/2500 of so called a great extent of scale reduction by a scanner or the like and then converting the read result into data are made to an object. A house recognizing trigger segment pair is selected, and when the segment pair exists, a view point is set up and visual segments are extracted. The presence of a distance between the end points of segments, i.e., the presence of gap is discriminated and whether a close rate found out from the total length of segments and the total length of gaps is more than a prescribed value or not is discriminated. Finally a house rule, i.e., that a closed graphic is formed of connected segments whose distance is less than a prescribed value, the number of segments is >=3, at least two segments have length more than a prescribed length, angles formed by the segments are 90 deg. or 270 deg. with a deviation within a prescribed value is discriminated and the discrimination is performed for >=3 samples of rule.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for recognizing a house figure from map line segment data, and particularly to a house / building figure area from line-divided map data when an electronic database such as a paper map is created. For extracting and recognizing.

[0002]

2. Description of the Related Art Conventionally, regarding the recognition of a house figure from line segment map data, a human is tracing a house part with a digitizer or the like, or a method of designating an end point sequence of line segments forming a house, or an automatic recognition device. In the case of, the mainstream method is to trace the end points of a line segment forming a closed figure in consideration of distance and direction.

[0003]

Among the conventional techniques, the method by which a person traces a house line segment on a map has a problem that it takes a lot of time and labor. In addition, the method of recognizing the house by tracking the distance and direction of the end points of the line segment is to locally check the position and direction of the line segment data and determine the line segment / direction to proceed to from the local state. Therefore, the house line segment data must be good quality data that is completely connected, and if the line segment data has poor quality such as breaks or blurs, it is recognized. In many cases, the programs did not work properly, and there was a problem that the recognition rate did not increase.

In view of the above-mentioned problems, the present invention adopts a global approach method of appearance, extracts line segments which are [visible] from a certain point, and checks whether those line segments satisfy house conditions. By doing this, it is possible to cover local conditions and recognize even houses that are not necessarily composed of completely connected line segments, thus significantly improving the recognition rate. The purpose is to

[0005]

A method of recognizing a house figure from map line segment data according to the present invention is the above-mentioned line segment data when extracting and recognizing a house figure from a set of line segment data representing a map. The line number that forms a right-angled corner is searched for from among the listed visible line segments as the starting point of the recognition process, and the number of corners that are almost perpendicular to the end point interval formed by the line segment row is calculated. It is recognized as a house figure based on.

Another feature of the present invention is that
A method of extracting and recognizing a house figure from a set of line segment data representing a map, in which each of the endpoints is two adjacent line segments within a specified distance, and at least one of them is greater than or equal to a specified value. And a line selected in the first process, which selects a line segment pair having a length of 1 and an angle formed by the line segments when each line segment is regarded as a straight line is close to 90 degrees. A point called a viewpoint is set near the midpoint of the hypotenuse of the pair of minutes, and line segments that are within a radius of at least 1/2 or more of the length of the hypotenuse from the viewpoint or that pass through the region are listed. Up, remove line segments with a length less than the specified length from those line segments, find the visible line segments from the viewpoint for the remaining line segments, and set the direction angle of each line segment to the specified direction. The data that is arranged in the following manner is created, and the end point coordinates of each visible line segment are also calculated. And the visible line segment data obtained in the second processing described above, the gap distance between the end points of the line segments that are sequentially adjacent in the predetermined direction is checked, and for the portion where the gap exceeds the specified value, the midpoint of the gap or If there is a visible line segment that is connected to the left and right line segments within the specified distance within the specified distance to the front (in the direction of the void) from the viewpoint, move the viewpoint to a point that is a minute distance to the left of the midpoint, and if there is The data inserted into the line segment sequence obtained up to that time is created, and if a void of a specified distance or more still occurs at this stage, the same process is repeated to obtain the third process and the third process. A fourth process for sequentially obtaining the maximum value of the interval between the line segment and the next line segment and the number of right-angled corners at which the angle between the line segment and the next line segment is approximately 90 degrees or approximately 270 degrees; If the number of right-angled corners obtained in the fourth process is 3 or more, One, the gap interval is a specified value or less of what has and a fifth processing recognizes the house candidate.

Another feature of the present invention is that when the number of right-angled corners obtained in the above-mentioned fourth treatment is 2 or more and the gap interval is larger than a specified value, a line segment is formed in the gap. Supplementally, there is a sixth process for recognizing a house candidate when the number of right-angled corners is 3 or more.

Another feature of the present invention is that the house candidates recognized in the fifth process are further checked by adding house conditions if necessary, and the house that satisfies the specified condition is selected. It has 7 processes.

Another feature of the present invention is that the number of right-angled corners is 4 as the additional house condition.
I am trying to be above.

Another feature of the present invention is that, as the additional house condition, the house is not recognized when the void distance is larger than a certain specified value.

[0011]

Since the present invention comprises the above-mentioned technical means, it becomes possible to globally recognize a house without being confused by a defect in a local condition, and the line segment is not necessarily completely connected. Since it becomes possible to satisfactorily recognize even a house that does not exist, the probability of recognizing a house part can be significantly improved.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for recognizing a house figure from line segment data representing a map of the present invention will be described below with reference to the accompanying drawings.

First, a schematic structure of a house figure recognition apparatus for carrying out the method of the present invention will be described. As shown in the configuration diagram of FIG. 1, the house figure recognition apparatus according to the present embodiment includes a line segment data input unit 1, a data processing unit 2, a program storage unit 3, an image data storage unit 4, an operation unit 5, and a display device. 6 and an output device 7.

The line segment data input unit 1 is provided to take in, for example, line segment data representing a map into the house figure recognition device, and is composed of, for example, a scanner. The data processing unit 2 is for performing a process of recognizing a house figure from a set of line segment data input from the line segment data input unit 1, and is configured by a CPU.

The program storage unit 3 is provided to store a program for the recognition processing performed by the data processing unit 2, and the image data storage unit 4 is loaded by the line segment data input unit 1. It is provided to store line segment data.

The operation unit 5 is composed of, for example, a keyboard, and is provided for the operator to input a command for causing the data processing unit 2 to perform a necessary process. . The display device 6 is for displaying the procedure and the result of the processing performed by the data processing unit 2, and is constituted by, for example, a CRT. Furthermore, the output device 7 is a device provided for outputting the result processed by the data processing unit 2 to a medium such as paper.

Next, the line segment map data of the present embodiment will be described. In the present embodiment, the target is line segment data which is converted into data after reading the city map data of 1/2500 or less, which is called a large scale, with a scanner or the like. In the following description, [point] is also treated as a line segment having a length of 0.

First, the problem is how to define / represent [house]. In the present embodiment, a house figure is defined as one that satisfies the following conditions. First, a house is a closed figure in which the distances between the end points of the line segments forming the house are connected at a specified value or less.

Secondly, the number of line segments constituting a house is at least 3 and at least 2 is at least the specified length.

Thirdly, when a corner which is an angle formed by adjacent line segments forming a house and has a deviation within a prescribed value of 90 degrees or 270 degrees is called a right-angled corner, there are at least three right-angled corners.

Fourthly, the interior of the house figure may be blank or have minute line segments of characters or symbols with a length equal to or less than a specified length. When these minute line segments are present, they are separated from the house-constituting line segment by a specified distance or more.

Fifth, even if the end point spacing between adjacent line segments is larger than the specified value, if there are only minute line segments corresponding to characters / symbols between them, both line segments intersect in extension. To consider.

Sixth, depending on the quality of the target map line segment data, a prescribed value is separately set, and even if the end point interval is larger than the initial prescribed value, if it is less than this value, relief may be performed.

In the house recognition method of this embodiment, when a line segment data set representing an urban area map is given, a pair of trigger line segments, which is a starting point for individual house recognition, is detected and a candidate house line segment sequence is detected. It is composed of three parts: the configuration and the house determination using the house conditions.

Next, an embodiment of a method for recognizing a house figure from map line segment data of the present invention will be described with reference to the accompanying drawings. 2 to 13 are flowcharts showing the procedure of the recognition method of the present embodiment, and FIGS. 14 to 18 are diagrams showing specific examples.

First, the overall processing procedure will be described with reference to the flowchart of FIG. When the process starts, step S
The map line segment data is read in 1 and then the house recognition condition is set in step S2. Next, in step S3, preprocessing / initialization of input data is performed.

Next, in step S4, a house recognition trigger line segment pair is selected, and in the next step S5, it is determined whether or not a corresponding line segment pair has been found. As a result of the above determination, if there is a corresponding line segment pair, the process proceeds to step S6, and the viewpoint is set and the visible line segment is extracted.

Next, in step S7, it is determined whether or not there is a void larger than the specified value, and in step S8, it is determined whether the closing rate is equal to or higher than the specified value. Then, if the result of these judgments is NO, the process jumps to step S12, and if the result is YES, the process advances to step S9.

In step S9, line segment tracing processing is performed by moving the viewpoint, and thereafter, in step S10, it is determined whether or not there is a void larger than the specified value, and step S11 is performed.
Check whether or not the closing rate is above the regulation. Then, as a result of the above determination, if there is no void more than the specified value or if the closing rate is more than the specified value, the process proceeds to step S12, and if not, the process returns to step S4. In steps S12 and S13, a determination is made according to a house rule described later, and the process ends.

Next, the input processing of map line segment data will be described with reference to FIG. When the process is started, the process of reading the start points, the end point coordinates, and the line width data of all the line segments into the input data file is performed in step S1, and then step S2.
At, a process of counting all line segment numbers is performed. Next, in step S3, x-, y-
The process of obtaining the maximum and minimum coordinate values is performed, and the process of inputting map line segment data ends.

Next, the setting process of the house recognition condition will be described with reference to FIG. In this case, first, the preprocessing condition is set in step S1, and then the house recognition condition setting process is performed in step S2. By the setting process of the house recognition condition, the minimum line segment length, the right angle determination allowable angle, the allowable gap length, the minimum circumference length, the minimum closing rate, the required number of right angle corners, etc. are set. Then, in step S3, post-processing conditions are set, and this processing ends.

Next, the preprocessing / initialization processing will be described with reference to FIG. First, in step S1, line segment end point connection processing is performed. Next, in step S2, the overlapping line segments are removed, and in step S3, registration in the rmd high-speed search file is performed. Then, in step S4, all line segments are unprocessed. The initialization process ends.

Next, the house recognition trigger line segment pair selection processing will be described with reference to FIG. First, in step S1, a selection process of a line segment having a specified length or more is performed. This process is a process of sequentially scanning unprocessed line segments and selecting a line segment having a length equal to or longer than a specified length. Next, in step S2, it is checked whether or not there is a corresponding line segment, and if not, this process is processed.

On the other hand, if there is, a neighborhood line segment listing process is performed in step S3. By this process,
All the line segments that are in the rectangular area with the specified width around the line segment are listed. After that, the orthogonal line segment is selected in step S4, and the presence or absence of the line segment orthogonal to the corresponding line segment within the allowable angle is checked.

Next, in step S5, it is checked whether or not there is an orthogonal line segment, and if not, the process ends. Also,
If there is, the process proceeds to step S6, the end points of the line segments orthogonal to the corresponding line segment are arranged counterclockwise, and the line segment number pair and the end point sequence data are output.

Next, the viewpoint setting and the visible line segment extraction processing will be described with reference to FIG. First, in step S1, a viewpoint and a search radius are set. First, using the coordinate data created in step S6 of the process of FIG. 6, a right-angled triangle is assumed and the coordinates of the right-angled point and the midpoint of the hypotenuse are determined. Next, the line segment from the right-angled point to the midpoint of the hypotenuse 1.0-1.
The viewpoint coordinates are set at the extension point of 9 times, and 0.6 to 2.0 times the length of the hypotenuse is set as the search radius.

Next, in step S2, a process of listing the line-of-sight segment is performed. This is done by setting the search radius as the radius with the viewpoint as the center, or by setting a square area having a length twice the search radius in the x- and y-coordinate directions, and setting a line segment that covers this area as much as possible. It is something to list.

After that, the visible line segment areas are obtained from the listed line segments in step S3 and arranged counterclockwise. The data is the line segment number, the direction angle of the start point and the end point, and the coordinate position data.

Then, in step S4, the total length of the line segments, the distance between the end points of the adjacent line segments and the total thereof are calculated, and if there is an interval of a prescribed length or more, the number is counted. . Next, in step S5, the closing rate = 100 * line segment total length / (line segment total length + interval total length) is calculated.

Next, the line segment search processing by moving the viewpoint will be described with reference to FIG. First, in step S1, a new viewpoint setting process is performed. This process retrieves the start and end coordinates of the last line segment in the line segment list that has been configured up to that point,
The new viewpoint is set at the position of the specified distance on the left side of the point where the start point → the end point is extended by the specified length.

Next, in step S2, the line segments of the area within the specified radius are listed from the new viewpoint, the visible line segments from the new viewpoint are obtained, and the lines are arranged counterclockwise. After that, a connection process is performed in step S3. This is to find the position of the last line segment that has been constructed so far from the visible line segment list arranged in the counterclockwise direction in the previous step S2, and calculate the end point spacing with the next line segment. Is.

Next, in step S4, it is checked whether or not the end point spacing is within a specified value. If the end point spacing is within the specified value, the process proceeds to step S5 to additionally register the line segment in the line segment list, and it is checked in step S6 whether or not the line segment is additionally registered. When additional registration is performed, the process returns to step S1. However, if it matches any one of the line segment lists configured up to now, it is determined that a loop is formed and the process proceeds to the next step S7.

In step S7, the sum of the lengths of the line segments, the distance between the end points of the adjacent line segments and the sum thereof are calculated,
If there is an interval with a specified length or more, count the number. After that, in step S8, the closing rate = 100 * line segment total length / (line segment total length + interval total length) is calculated, and the process ends.

Next, the calculation process of the house feature amount will be described with reference to FIG. First, in step S1, the process of obtaining data such as the number of line segments, start / end point coordinates, closing rate, and total length is completed. In the next step S2, the obtained line segment sequence is regarded as a closed figure, the interior angle of the adjacent side is calculated, and if it is 90 degrees or 270 degrees within the allowable deviation range, it is regarded as a right angle corner and the number is counted. Up.

Then, in step S3, the processing for obtaining the squareness is performed. This is the squareness = 100 * 16 *
The area / (perimeter squared) formula is used, and the non-symbol ratio is calculated by the formula: non-symbol ratio = 100 * total length of line segment / (number of line segments * symbol line segment reference length) -50 To be done.

Next, the determination process according to the house rule will be described with reference to FIG. This is a determination based on the house determination feature amount (step S1) and the house accuracy (step S2). In step S1, when the orthogonal corner is n, the probability p1 of being a house is p1 = 2.
5 * n% (100 or more possible), closing rate p2 (max10
0), non-symbol rate p3 (max100), and squareness p4 (max100). Also,
In step S2, p = p1 * p2 * p3 * p4, p>
If = p0, the house is set (p0, for example, 75%).

Next, the house recognition process will be described with reference to FIG. First, in step S1, at least one side has a specified length or more, and two sides that are substantially right angles are selected.
Next, the presence or absence of the corresponding line segment is determined in step S2, and if there is, a viewpoint is set near the midpoint of the hypotenuse in step S3, and a line in a radius region of at least ½ length of the hypotenuse or more is set. List minutes.

Next, in step S4, it is searched whether or not the line segment has three or more sides. If there is, the visible line segment from the viewpoint is obtained in step S5, and the process of arranging the line segments counterclockwise is performed. . Next, in step S6, it is searched whether or not there are three or more sides. If there is, the process proceeds to step S7, and it is searched whether or not there is a void larger than the specified size between the two adjacent sides.

As a result of the above retrieval, in step S8, the viewpoint is reset inside the vicinity of the center of the void, and the visible line segments are listed. Then, in step S9, the line segment row up to that point is connected. By this processing, it is checked whether or not these line segment sequences are closed, and if they are closed, the process proceeds to step S11.
If it is not closed, the process returns to step S8 and the above-described processing is repeated. Furthermore, if it cannot be closed, the process returns.

In step S11, 90 ° or 270
Check and count the number of corners that make up °. And 3
If there are more than one, the process proceeds to step S13 and the house candidate is recognized. In the case of two, the process proceeds to step S14 and is recognized as an incomplete house candidate. On the other hand, when the number is one or less, the process proceeds to step S15 and it is recognized that the house is not a house candidate.

A specific example for realizing the above method will be described below. First, selection of a pair of trigger line segments and setting of a viewpoint will be described. If the determination of the pair of orthogonal line segments that triggers house recognition is an interactive process, if a human specifies one point inside the house figure with a mouse, etc., that point can be directly set as the viewpoint. , It is not necessary to search for a pair of orthogonal line segments. Here, a method for finding a pair of orthogonal line segments will be described for the purpose of processing mainly based on automatic recognition. An explanatory diagram of this method will be described with reference to FIG.

First, in step S1, unprocessed 1
Perform the process of selecting a line segment. An unprocessed line segment having a specified length or more is selected from the given line segment data set and is set as the first line segment A (step S2). If there is no unprocessed line segment, the process ends.

Next, the neighboring line segments having the end points within the specified distance range from the end points of the line segment are listed up. The presence or absence of such a line segment is searched for in step S5, and if there is no corresponding line segment, the first line segment A cannot be a trigger and the process returns to step S1. Next, in step S6, the interior angles of the first line segment A and the second line segment B are calculated, and in step S7, the interior angle that can be regarded as 90 degrees within the allowable deviation is searched.

Then, if there is a corresponding line segment, a process of making it the second line segment B is performed in step S8. Next, proceeding to step S9, assuming that the selected line segment pair is A → B without loss of generality, the first and second line segments A and B are assumed to be two sides sandwiching a right angle of a right triangle. Then, the midpoint of the hypotenuse or one point in the vicinity thereof is determined, and this is set as the viewpoint.

Next, the creation of the house candidate line segment will be described more specifically. First of all, focusing on the viewpoint,
After listing the line segments that are included in a region whose radius is at least 1/2 of the hypotenuse of the right triangle or that pass through that region, and remove minute line segments that are less than the specified length,
The visible line segment portion that can be seen from the viewpoint is obtained from that, and line segment data D1 arranged counterclockwise is created. FIG.
An example in the case of a convex figure house is shown in (a) and (b).

Secondly, the distance between the end points of adjacent line segments of the first line segment row is calculated, and if the distance is equal to or greater than a specified value, the following processing is performed. That is, as shown in FIG. 14, it is assumed that the line segment is examined counterclockwise and there is a distance between the end points that is equal to or greater than the specified value between the line segment en and the line segment en + 1. At this time, from the point of d1 on the extension of the line segment en to the left HV
A new viewpoint is set at a point of the distance of IEW, the line segment en is viewed at the right end, and a portion where the end point distance is equal to or more than a specified value is looked into from the viewpoint, and a visible line segment is extracted.

Next, it is checked whether or not there are visible line segments such as the line segment en and the line segment en + 1 that are connected within the specified distance at the end points of the line segment of the data string D1. Then, if such a line segment sequence exists, it is additionally inserted at the end point portion of the data sequence D1.

If there is a break point more than the specified distance in this additional line segment sequence, the same processing is repeatedly applied. However, in the extraction of visible line segments, minute line segments with a specified length or less are excluded. If the distance between the end points is larger than the specified value, and there are multiple minute line segments between the end points, and both line segments are on the same straight line or are orthogonal, a line segment between both end points To add. If these conditions are not satisfied, the process is stopped as an open figure.

Thirdly, the end points of the line segment row formed in the above-mentioned second step are covered with a rectangle with a specified value width. By this processing, the same line segment divided into a plurality of lines can be restored to one line segment. This line segment sequence extracts not only a convex figure but also a figure including a concave section, and this is a house candidate line segment sequence. FIG. 15 shows an explanatory view of the rectangular coating.

Next, the determination of the house figure will be described.
In this embodiment, the house determination is performed according to the following procedure. That is, (1) the area of the house candidate closed figure and the interior angle of the adjacent line segment are calculated. If the interior angle is 90 degrees or 270 degrees within the allowable deviation, the sum of the numbers is calculated as a right-angled corner.

Next, (2) if the area is within the specified value range and the number of right-angled corners is 3 or more, it is determined to be a house. further,
(3) If the area is within the specified range and the number of right-angled corners is 2 or less, it is determined that the house is incomplete. (4) In other cases, it is not considered as a house.

16 (a), (b) and (c) show examples of house determination. FIG. 16 (a) has four or more right angles,
In addition, it shows an example of a complete house in which the surroundings are all closed. Moreover, although (b) is incomplete, the above (1)-
Since the condition (3) is satisfied, it is recognized as an incomplete house. Further, although (c) has three or more right angles, it is recognized that it is not a house because the unclosed portion is large.

Program the above road recognition method to 1
The following is the result of reading the / 2500 city map at a resolution of 16 points / mm and performing recognition processing on the vectorized data by the method described above.

An example of the recognition result from one viewpoint is shown in FIGS. FIG. 17 shows the original data of a relatively small area, and FIG. 18 shows the house recognition result from FIG. As is clear from these figures, house figures that do not necessarily consist of right-angled corners are often extracted.

[0065]

As described above, according to the present invention, it is possible to globally recognize a house without being confused by a local defect. Therefore, even a house that does not necessarily consist of completely connected line segments can be recognized as a good house, and the probability of recognizing a house part from the set of line segment data that represents a map can be calculated. It can be greatly improved.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an example of a house area recognition device that implements the method of the present invention.

FIG. 2 is a flowchart showing an overall processing procedure of a method for recognizing a house figure from map line segment data according to the present invention.

FIG. 3 is a flowchart showing a procedure for inputting map line segment data.

FIG. 4 is a flowchart showing a procedure for setting house recognition processing conditions.

FIG. 5 is a flowchart showing a pre-processing / initialization processing procedure.

FIG. 6 is a flowchart showing a procedure for selecting a house recognition trigger line segment pair.

FIG. 7 is a flowchart showing a procedure for setting a viewpoint and extracting a visible line segment.

FIG. 8 is a flowchart showing a procedure of line segment search processing by moving a viewpoint.

FIG. 9 is a flowchart showing a calculation process of a house feature amount.

FIG. 10 is a flowchart showing a determination process based on a house rule.

FIG. 11 is a flowchart showing a processing procedure for house recognition.

FIG. 12 is a flowchart showing a search procedure for a pair of orthogonal line segments.

FIG. 13 is a diagram showing an example of a convex house.

FIG. 14 is a diagram illustrating a method of setting a viewpoint on a concave portion.

FIG. 15 is a diagram illustrating rectangular coating.

FIG. 16 is a diagram showing an example of a house determination.

FIG. 17 is a diagram showing an example of original data.

18 is a diagram showing an example of a result of performing house recognition from the data of FIG.

[Explanation of symbols]

 1 line segment data input unit 2 data processing unit 3 program storage unit 4 image data storage unit 5 operation unit 6 display device 7 output device

Claims (6)

[Claims] 1. When extracting and recognizing a house figure from a set of line segment data representing a map, a pair of line segments forming a right-angled corner is searched from the line segment data to be used as a starting point of recognition processing. Map line segment data characterized by listing line-of-sight segments, determining the number of corners that are almost perpendicular to the end point interval formed by the line segment sequence, and recognizing the figure as a house based on the number of corners determined above How to recognize a house shape from a map. 2. A method for extracting and recognizing a house figure from a set of line segment data representing a map, wherein each of the end points is two adjacent line segments within a prescribed distance, and at least One has a length greater than the specified value,
And when each line segment is regarded as a straight line, the angle formed by the line segment is 90
A first process of selecting a line segment pair close to a degree, and a point called a viewpoint is set near the midpoint of the hypotenuse of the line segment pair selected in the first process, and at least the length of the hypotenuse from the viewpoint is set. The line segments that are within the radius of 1/2 or more of or that pass through the region are listed, and the line segments with the specified length or less are removed from those line segments. A visible line segment from the viewpoint is obtained, data is arranged so that the direction angle of each line segment is around a predetermined direction, and the end point coordinates of each obtained visible line segment are also obtained.
And the visible line segment data obtained in the above second process, the void distance between the end points of the line segments that are successively adjacent in the predetermined direction is checked. Alternatively, move the viewpoint to a point a minute distance to the left of the midpoint, and check the presence of a visible line segment that connects within the specified distance to the left and right line segments of that void, from the viewpoint to the front (air gap direction). For example, when the data inserted into the line segment sequence obtained up to that time is created, and at this stage, if a void of the specified distance or more still occurs, the same processing is repeated in the third processing and the third processing described above. With respect to the obtained line segment sequence, a fourth process of sequentially obtaining the maximum value of the interval with the next line segment and the number of right-angled corners whose angle between the next line segment and the next line segment is approximately 90 degrees or approximately 270 degrees. , The number of right-angled corners obtained in the fourth processing is 3 or more. In,
A method of recognizing a house figure from map line segment data, which comprises a fifth process of recognizing a space candidate having a gap value of a predetermined value or less as a house candidate. 3. When the number of right-angled corners obtained in the fourth processing is 2 or more and the space between voids is larger than a specified value, the number of right-angled corners becomes 3 or more by supplementing the space with a line segment. The method for recognizing a house figure from map line segment data according to claim 2, further comprising a sixth process of recognizing the house candidate. 4. A seventh process for checking the house candidate recognized in the fifth process by further adding a house condition, if necessary, and selecting a house that satisfies a specified condition as a house. A method for recognizing a house figure from the map line segment data according to item 2. 5. The method of recognizing a house figure from map line segment data according to claim 4, wherein the number of right-angled corners is 4 or more as the condition of the house to be added. 6. The house figure is recognized from the map line segment data according to claim 4, characterized in that, as the additional house condition, the house distance is not recognized as a house if the void distance is larger than a certain specified value. Method.