JPS6174075A - Segment pattern preparing system - Google Patents

Abstract

PURPOSE:To compress segment pattern data which become necessary in a high speed polygonal approximation by sharing data of the common part when plural segment patterns exist. CONSTITUTION:A segment processing part 107 stores the said respective peak coordinates, in which the area is polygonally approximate, to a segment information memory 108 successively. When a dictionary pattern is registered, the segment information is shifted to a dictionary pattern memory 109. A common area extracting part 112 converts coordinates at an input pattern side for the pattern inputted at this time (the pattern stored in the segment information memory 108) and the dictionary pattern (the pattern stored in the dictionary pattern memory 109) so that characteristics of both patterns can be coincident. Common areas of both pattern (product pattern) are obtained as a peak row in the same manner and recorded in a common area memory 113.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a line segment pattern creation method used in a component recognition device button by image processing. The present invention relates to a line segment pattern creation method suitable for performing a process of approximating a line segment (contour point sequence) at high speed.

[Background of the invention]

Conventionally, it has been proposed to perform component recognition by approximating a component pattern into a polygon for the purpose of high-speed and flexible image processing. At this time, all line segment patterns are stored and the sides of the polygon are extracted at high speed by comparing the trunk point sequence of the part with this. Regarding this, for example, see Proceedings of the National Conference of the Institute of Electrical Engineers of Japan, published on May 10, 1988 [15], paper number 1540K, ``Part recognition by line segmentation pattern matching.''

However, in these conventional methods, the line segment pattern data used in the line segmentation process includes all straight line patterns within a certain length, so the memory size is large (which makes it difficult to make the device smaller and more economical). This was a problem.

[Purpose of the invention]

The purpose of the present invention is to solve the above-mentioned conventional problems, compress line segment pattern data necessary for high-speed polygon approximation, and provide a line segment pattern creation method that can be executed even by a small-scale one-man processing device. Our goal is to provide the following.

[Summary of the invention]

When there are multiple line segment patterns, even if the shapes are different as a whole, some partial pressures may be common. Therefore, in the present invention, by sharing the data in this common portion, the data is reduced and the required memory size is reduced.

[Embodiments of the invention]

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of an image processing apparatus according to the present invention.

Here, 100 is a control section that controls the flow of required processing for each section described below.

Reference numerals 101 to 104 relate to preprocessing, and 101 is a human control unit, 102 is an image memo IJ, 105 is a preprocessing unit, and 104 is an area information memory. 105 and 106 are related to contour extraction processing; 105 is a contour extraction unit;
106 is a contour memory. 107 to 109 are related to line differentiation processing, 107 is a line differentiation processing unit, 115 is a
Line segment pattern memory 108 is line segmentation information memory 1
09 is a dictionary pattern memory. 110 and 111 are related to region feature extraction processing, and IjOf'? , , the region feature extraction unit 111 is a region feature memory. 112
114 are related to pattern matching processing, and 11
2 is a common area extraction unit; 115 is a common area memory; 11
4 is a recognition processing section. Note that the one-man processing device 4 is composed of a processor, a memory, and hard circuits such as an interface circuit.

First, image data input from a television camera (not shown) is digitized by the input control unit 101 and binarized using a predetermined threshold value, and then inputted to the image memory 102 as grayscale image data and binary image data. be remembered.

The image data stored in the first love memo 1JI02 is first determined by the busy pre-processing unit 105 as to which area on the image is to be the target object area, and then the position information of the representative point of that area is stored in the area information memo 1. Stored on month 04.

Next, the contour extraction unit 105 determines a search start point based on the position information of the representative point, sequentially searches for contour pixels in the specified area on the image, and stores, for example, the center coordinate value in the contour memory 106. let

Based on the center coordinate value, the line segmentation processing unit 107 traces the line of points representing the center of the contour pixel point by point and compares it with the line segment pattern data preset in the line segment pattern memory port 5 to segment the line segment. The end points of the line segment,
That is, the coordinates of each vertex obtained by approximating the area to a polygon are sequentially stored in the line segmentation information memory 108. The line division information is transferred to the dictionary pattern memory 109 when registering the dictionary pattern.

Further, the region feature extraction unit 110 obtains features of the specified region (for example, area, N center, principal axis of inertia, maximum length 1 perimeter, etc.) from the contents of the line segmentation information memory 108, and
Store it in 1. When adding a dictionary pattern, its characteristic information is stored in the dictionary pattern memory 109.

Further, the common area extraction unit 112 extracts both the currently input pattern (stored in the line segmentation information memory 108) and the dictionary pattern (stored in the dictionary pattern memory 109). After converting the coordinates on the input cover turn side so that the features match (for example, so that the center of gravity coordinates match or the principal axis of inertia directions match), convert the common area (product shape) of both patterns in the same way. is obtained as a vertex sequence and recorded in the common area memo 1JII5.

Finally, the recognition processing unit 114 calculates the area of the common area (product figure), calculates the ratio of this to the area of the dictionary pattern or the input pattern, and if the value exceeds a predetermined value, both patterns are -iL, and outputs the recognition result.

If the two patterns do not match and there are multiple dictionary patterns, the second pattern. The matching process with the fifth dictionary pattern is repeated in the same way. Note that the recognition processing unit 114 calculates the position of the input cover turn from the relative position before conversion of both matching patterns.

Posture information is also output.

In the above-mentioned series of 1iij (image processing), a method of creating a line segment pattern used in the line segment pattern memory 115, which is the essential part of the present invention, and processing using the same will be described next.

The polygonal approximation performed by the line segmentation processing unit 107 is a process in which the outline of the target image area is traced, and when a series of continuous outline points can be regarded as a straight line, both end points thereof are set as the vertices of the polygon.

As shown in Figure 2, there are two ways to think about contour lines: (A) and (B), where the center of a contour pixel is the contour line, and (C''), where the boundary line between area pixels and background pixels is the contour line. ,
There are two ways (D), and in addition, there are 4 connections (B), which considers only vertical and horizontal adjacency for the adjacency state of continuous contour points.
(D), 8-connection (A) that also takes into account diagonal adjacency, (
There are two options (C). Therefore, for contour tracing, (A)
There are a total of four methods O (D). However, among these, the boundary 8 connection (C) is meaningless because it does not uniquely correspond to the pixel value.

In each case, the direction of tracking is
), (b), and (cL). Hereinafter, a case will be described in which a contour line is traced using boundary 4 connection (D).

An example of a contour point sequence is shown in FIG. Each contour point has its own direction of movement, as indicated by the arrow in the figure, and the shape of the contour line is represented by a column of this direction of movement. If a restriction is placed on the length, the sequence of points that are considered to be line segments under certain conditions becomes a finite number of sequences in the advancing direction. As the condition for considering a line segment, for example, a maximum permissible distance from the center line is determined, and a line segment is defined as a point sequence in which all points satisfy this value. Since the number and length of line segment patterns are finite, all line segment patterns can be created artificially.

FIG. 4 shows an example of a representation format of a line segment pattern and polygon approximation using the same.

The line segment pattern is 5K as shown in the 4th @(α).

It is represented by a 5×1 two-dimensional array, and this is stored in the memory 115 as pattern data. The second subscript value of the array corresponds to the number of points on the line segment pattern represented by the dot sequence, and the first subscript value corresponds to the three directions of travel at each point, namely, left turn, *forward and right turn. The contents of the array element are the number, ie, the pointer, of the next point when the direction of movement is determined at a certain point. For example, when the direction of travel is determined to be a right turn at the start point addition', the number of the next point when proceeding in that direction is indicated as @187'. However, since there is a restriction that the series of points can be regarded as a straight line, all six traveling directions are not always allowed at each point. An end mark END is written in the element corresponding to the prohibited direction, and when this is detected while tracing the contour, the detected point is regarded as the end point of the line segment, and this is registered as the vertex of the polygon. do. In FIG. 4(b), the starting point 1 of the line segment pattern is at the starting point S of the line segment.
Start tracing from "o", detect the end mark END at point E, set this point °333" as the vertex of the polygon, return the line segment pattern pointer to the starting point "o2, and start tracing a new line segment. The number of each point inside is the point number of the line segment pattern.

Next, the plane of polygon approximation using this line segment pattern will be explained using FIG. 4 as an example using FIG. 5. In polygonal approximation, contour points are traced one by one, and at the same time, a line segment pattern is traced by selecting a pointer according to the direction of movement at each point. First, the tracing start point, for example, the eight points in FIG. 4(h), are registered as the first vertices of the polygon, and the line segment pattern pointer is initialized to "Om" (step 5a1).

Thereafter, the processes of steps 501 to 507 are repeated.

First, the traveling direction at the tracking start point S is determined (step 502). Here, a method for determining the traveling direction will be explained using FIG. 5. In FIG. 5, it is assumed that the pixel person is included in the component area. First, it is checked whether pixel B is included in the component area, and if it is not included, the traveling direction is set to be a right turn. Next, check whether pixel B is included in the component area, and if it is not included, move forward in the direction of movement.
If included, turn left. There are various methods for determining whether a pixel is included in a component area. For example, in the case of a binary image, determination can be made based on whether the value read from the contour memory 106 is 11 or 10'.

Next, referring to the slot pointed to by the pointer of the line segment pattern, the next pointer is selected according to the direction of movement (step 50S). For example, in the case of FIG. 4, the area of pointer 0'' is referred to, and in this case, since the direction of travel is a right turn as shown in FIG. 4(A), the next pointer @187° is selected.

If the next pointer is the end mark END (step 504), the contour point at that point is newly registered in the line division information memory 108 as the vertex of the polygon, and the pointer is returned to @02 (step 505). .

If not, proceed to the next contour point according to the traveling direction (step 5a6) and check the end condition, for example, whether or not the contour has been circled once and returned to the tracing start point S.Step 507
), steps 502 to 507 are repeated until the termination condition Di is satisfied. For example, Fig. 4(b)
When tracing the contour line according to the line segment pattern data shown in FIG. 4 (α), it starts from the pointer 10'' and advances to the next pointer ``187'' according to the direction of travel (right turn).

Thereafter, the next contour tracing is started, the direction of travel (straight ahead) is determined, and the pointer 515' is selected. Thereafter, tracking is performed in the same way, and when a pointer indicating the end mark END is detected, the pointer "3551" at that time is registered as the vertex.The coordinate position at this time is calculated based on the coordinates of the starting point and stored in the memory 108. to be memorized.

However, as shown in Figure 4 (α), the line segment pattern, N, is a pair of axially symmetrical lines, and the first turning point is only one with a right turn (or left turn). After the first corner point O (appears) on the contour point sequence, when that corner point is a left turn (right turn), the left and right of the line segment pattern is reversed and used.

Since the line segment pattern shown in FIG. 4(α) is constant data, it can be created offline and later loaded with the program. Therefore, there is no need to perform high-speed processing during creation. In addition, since the contents of the line segment pattern are unrelated to the processing procedure of polygon approximation, operations such as adjusting the roughness of polygon approximation can be programmed by replacing the line segment patterns created using various straight line criteria. This can be easily done without changing the hardware or hardware.

First, the procedure for creating a line segment pattern as shown in FIG. 4 (α) will be explained. The line segment pattern is represented by a 5X array. The second subscript corresponds to each point on the line segment pattern, and the first subscript corresponds to each point in the traveling direction. The contents include the direction of movement at each point pressure, O) busy at a fixed time, a pointer indicating the next point, or an end mark END. Therefore, a line segment pattern is a set of slots in which the permissible traveling directions at each point are written on the condition that the line is a straight line.

To create a line segment pattern, at each point on the line segment pattern, add new points according to the five directions of progression,
It is determined whether the point sequence obtained by adding this is a straight line. If it is a straight line, an end mark is created, and if it is a straight line, a new slot corresponding to the new point is created, the number is written as a pointer, and the same process is repeated for the new point.

The above steps are multiplied by i until the length of the line segment reaches the allowable value.

This basic procedure can be carried out using a method known as the vertical method in the state-space search method. (This method is described, for example, on pages 55-56 of "Artificial Intelligence" published by Corona Publishing on April 15, 1972.

The above is the basic idea of creating a line segment pattern. The flow of processing will be explained step by step using FIG. 6 below.

The line segment pattern is represented by an array T. Let C be a counter that indicates how far a line segment pattern is used. Also,
In order to compress the axially symmetrical pattern, one-point sequence O'--by setting a flag to indicate whether it has already turned to the right or left, or whether it is continuing to go only straight, with an on/off setting, and by restricting the direction of the first bend to one side. Creates an axially symmetrical pattern on only one side, compressing the data in half.

First, the point sequence and counter are initialized as shown in FIG. 6(α)K. In the point sequence, the first two points are fixed to, for example, (0,0) and (+10). As a result, the line segment pattern is 4
Data is compressed using only one rotationally symmetric pattern.

Leave the flag turned off. Also, clear the counter to 0.

(Step 602
). The point sequence extension process further includes the next point sequence extension process, and the extension is repeated one after another until the permissible length is reached. The details of the point sequence extension process will be explained with reference to FIG. 6(h)K. Note that this Fig. 6 (h) is Fig. 6 (-)
FIG. 6 is a diagram showing a subroutine of step 602 of FIG.

First, as shown in Figure 6 (h), the point to be processed is designated by P. This pointer P is a local variable within the point sequence extension process, and is fixed to the value of C at the beginning of the point sequence extension process. (Step 611).The value of C itself changes during processing.Next, the counter and the point sequence length are incremented and a new point is added (Steps 612 and 615) assuming that the direction of travel is a left turn. That is, pointer entry processing (step 6
14). This pointer entry process is shown in Figure 6 (C).
As shown in FIG. 3, pointer entry processing is performed only when the flag is on (steps 621 to 625), and when it is off, an end mark END is written (step 627). As a result, the direction of the first turning point is limited to the right. Next, the direction of travel is set to go straight, and this time pointer entry processing is performed regardless of the e-flag (step 615). Next, the direction of travel is set to right, and the new A point is added, a flag is turned on, and pointer entry processing is performed (step 616).That is, the flag is turned on after a right turn is made at least once.

Next, regarding each pointer entry process in FIG. 6(h),
This will be explained in detail according to figure (C). This FIG. 6(C) corresponds to steps 614 and 615 in FIG. 6(A).

616 is a diagram showing each subroutine. First, a point sequence is extended (step 622), and it is determined whether the length of the point sequence exceeds an allowable value (step 625). As a result, if it has exceeded, an end mark is written (step 627), and if it has not exceeded, it is determined whether the point sequence to which the new point has been added is a straight line (step 625). If it is not a straight line, an end mark is written (step 627). Judgment base $+S Any value is acceptable as required. For example, a method is used in which the center helix of a point sequence is drawn using the least squares method and a maximum allowable value is set for the distance from this point to each point. If it is a straight line, increment the counter C to create a new slot and write this counter value as a pointer (step 625).
The length of the point sequence is incremented by performing the point sequence extension process shown in (step 626). In this way, the fourth
A line segment turn as shown in figure (α) is created. That is, pointer entry processing is performed by incrementing a counter from pointer @0', and if the first turning point is on the left, an end mark is written, and the pointer entry processing for straight travel is increased by vc to the allowable value.

Thereafter, pointer entry processing for a right turn and then a left turn is performed within the allowable value.

The line segment pattern created in this way includes all point sequences that are considered to be straight lines under a certain standard, and does not include double point sequences of the same shape. As shown in α) and (A), there are parts C and D of the same shape. When a part of the same shape includes the end point of a line segment, the point sequence data corresponding to that part will have double equivalent data, and by representing one of them and erasing the others, Data can be compressed.

The procedure will be explained using FIG. Since the portion to be erased always includes the end point of the line segment, that is, the slot containing the end mark in all three directions, unnecessary portions are erased going backwards from this point.

First, as shown in FIG. 8(α), K and the line segment pattern array are scanned to find the end point, and erasure marks are written in all but the first point (■) (step 801). Next, the line segment pattern array is scanned, and a series integration process is performed in which all but one point containing the symbol ■ is deleted (step 802).

The point sequence integration process will be explained using FIG. 8(h). First, specify the BOI/RIJ to be explored. Then, the next target construction is cleared (step 811). Next, the line segment pattern is scanned to find a point including J, and a point sequence erasing process is performed in which all points except the first point I are erased and written with i-ri (step 812). Next, if point I exists (step 815), a similar point sequence integration process is performed for point I (step 814).

Finally, erase all slots marked with erasure marks, back up the data, and exit (step 805).
.

By performing the following processing, the data of the common part between different line segment patterns as a whole, that is, the partial sequence in the advancing direction, is shared [Thus, the amount of necessary data is reduced and the line segment pattern memory 115 is compressed. I can do something.

〔Effect of the invention〕

As described above, according to the present invention, while taking advantage of polygon approximation processing in an image processing device, it is possible to significantly reduce the required memory size and achieve miniaturization and economicalization.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the image processing device used in the present invention, FIG. Figure 4 is a diagram showing an example of a contour line, Figure 4 is a diagram showing an example of line segment pattern expression format and polygon approximation, Figure 5 is a flowchart showing the procedure for polygon approximation, and Figure 6 is a line segment pattern creation process. 7 is a diagram showing an example of a line segment pattern that has a common part, and FIG. 8 is a flowchart of removing the common part of the line segment pattern (
It is a flowchart of data compression processing. 101... Input control unit, 102... Image memory, 1
05... Preprocessing unit, 104... Area information memory, 10
5...Contour extraction unit, 106...Contour memory, 107
...Line differentiation processing unit, 108...Line differentiation report mail 41J,
+09...Dictionary pattern memory, 110...Region feature extraction unit, 111...Region feature memory, 112...
Common area extraction unit, mouth 5... common area memory, mouth 4...
- Recognition processing unit, 100...control unit, 115...line segment pattern memory.

Claims (1)

[Claims] Polygon approximation of the part image is performed by comparing the contour point sequence of the part image with data stored in a line segment pattern memory that stores various line segment patterns required for line segment approximation, and the vertex coordinates of the polygon are calculated. In an image processing device that performs recognition processing based on columns, a partial sequence of common contour points among the various line segment patterns is shared and stored in the line segment pattern memory. Minute pattern creation method.