JPS59140589A - Outline extracting device - Google Patents

Abstract

PURPOSE:To improve the speed of processing of an outline extracting device by using a linear operation filter, and converting picture information of a two- dimensional structure into a linear structure in the initial stage of outline extracting process. CONSTITUTION:Edge detecting sections 21, 22 applies raster scanning horizontally and vertically to the output of a noise elimination section 1 by differential filters of 3X1 and 1X3 respectively. A group of points obtained by extracting the first point at which the temperature change attains maximum one by one for each scanning line, and a group of points obtained by extracting the last point at which temperature change attains minimum one by one are obtained as data of linear disposition. Picture information of two-dimension structure is converted to the linear structure by edge detecting sections 21, 22, and the subsequent processing can be performed by the linear processing at a high speed. Thus, an output extracting device for rectangle having increased processing speed can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field of the Invention The present invention relates to a contour extraction device for a rectangular object that achieves high processing speed through -dimensional processing.

(B) Background of the technology One application field of pattern recognition technology is a system for automatically sorting documents or books, etc. In such systems, the sorting is usually carried out by sorting the target items while transporting them on a conveyor or the like. Sorting is done based on the recognition results of symbols, characters, patterns, etc. described in
As part of the recognition process, a so-called "cutout" operation is often performed in which the outline of the recognition target is recognized in advance and the recognition target is extracted.

(C) Prior Art and Problems In conventional contour extraction devices for rectangular objects, a two-dimensional calculation filter is used to scan a rectangular object in raster directions, and the method is used to determine whether the magnitude of the density change rate at each pixel point or the density change rate is the maximum. Contour extraction processing was performed after determining both the direction in which the density change rate is maximum or the direction in which the density change rate is maximum and the direction in which the density change rate is maximum as a two-dimensional array.

Therefore, the contour extraction process takes a long time and requires a large-capacity register, making it difficult to put it into practical use.

α)) Purpose of the Invention The present invention improves the processing speed of a contour extraction device by converting image information of a two-dimensional structure into a one-dimensional structure at the initial stage of contour extraction processing using a -dimensional calculation filter. With the goal.

The contour extraction device according to the present invention scans a digital image of a rectangular object using a -dimensional calculation filter, and extracts points at which the density change rate is maximum one by one for each scanning line. an edge detection unit that extracts points at which the concentration change rate is minimum one by one and obtains each fish school as one-dimensional array data; and an edge detection unit that obtains each fish school as one-dimensional array data, and an adjacent distance of each fish school obtained by the edge detection unit is less than or equal to a predetermined value. a connecting line generation unit that obtains a connecting line by connecting the points in the order of scanning lines, and extracting the longest connecting line from the connecting lines obtained by the connecting line generating unit and linearly approximating the longest connecting line. An edge candidate generation unit that obtains side candidates of the rectangular object and a connection line that has a distance smaller than a predetermined value from the straight line obtained by the edge candidate generation unit among the connection lines obtained by the connection line generation unit. a side element extraction unit, a side slope candidate detection unit that linearly approximates all the connecting lines obtained by the side element extraction unit and detects the slope of the straight line, and a rectangle as described above obtained by the side slope candidate detection unit. a slope detection section that calculates the slope of the rectangular object from slope candidates for each side of the object; The present invention includes a side detection unit that detects sides, and extracts the outline of the rectangular object based on the slope of the rectangular object obtained after obtaining slope candidates for each side of the rectangular object.

In other words, focusing on the fact that the contour to be extracted consists of a total of four line segments, two of which are orthogonal to each other,
The processing is simplified by applying a -dimensional calculation filter, that is, a one-dimensional window calculation in two mutually orthogonal directions, and detecting the edges where the density change rate is maximum or minimum, and the 4-dimensional calculation filter obtained in this way The edges of the set are made to correspond to the four sets of contours, or sides, to be extracted, and the elements of each side are extracted to detect the slope of each side. This processing process requires only a small amount of information to be handled, so the processing is simple and can be performed at high speed. Finally, the slope of the object is determined by weighting the slope of each side and calculating the average 4-, and when combined with the information about the elements of each side obtained earlier, the detection of the four sides, that is, the extraction of the contour, is made very quickly. It is something that is done on time.

■Examples of the Invention The gist of the present invention will now be illustrated by way of examples.

The figure shows a system block diagram of an embodiment of the present invention, in which 1 is a noise removal unit that uses a median filter to remove noise components in an input digital image obtained by observing a rectangular object, and 21 is a noise removal unit of the noise removal unit 1. The output is scanned in the horizontal direction by a 3×1 differential filter, and the first point where the rate of change in concentration is the maximum is extracted one by one for each scanning line, resulting in a school of fish and the rate of change in concentration is the minimum. An edge detection section extracts the last points one by one and obtains each school of fish as data in a one-dimensional array. 22 performs a rask scan of the output of the noise removal section 1 in the vertical direction using a 1×3 differential filter. The point group obtained by extracting the first point at which the concentration change rate is extremely thick for each line, and the fish group obtained by extracting the last point at which the concentration change rate is minimum one by one, are each linearly Edge detection unit obtained as data of original array, 31 and 32
is the connecting line on the left by connecting the points of the two groups of fish obtained by the edge detection unit 21 in which the adjacent distance is less than or equal to the scanning line pitch ρ and the number of constituent points is 5 or more in the order of the scanning line. 33 and 34 respectively generate upper connections from the two groups of fish obtained by the edge detection unit 22 in the same manner as described above. A third connecting line generating section that generates the line, a fourth connecting line generating section that generates the lower connecting line, 41, 42, 43, and 44 are the first connecting line generating section 31 and the second connecting line generating section, respectively. The sides of the rectangular object are extracted by extracting the longest connecting line from among the connecting lines obtained by the unit 32, the third connecting line generating unit 33, and the fourth connecting line generating unit 34, and linearly approximating the longest connecting lines, respectively. A first edge candidate generation unit, a second edge candidate generation unit, a third edge candidate generation unit, and a fourth edge candidate generation unit, 51, 52, 53, and 5, which obtain candidates for
4 are the first sides of the connecting lines obtained by the first connecting line generating section 31, the second connecting line generating section 32, the third connecting line generating section 33, and the fourth connecting line generating section 34, respectively. A connecting line whose distance from the straight line obtained by the candidate generation unit 41, the second edge candidate generation unit 42, the third edge candidate generation unit 43, and the fourth edge candidate generation unit 44 is smaller than a predetermined value is extracted. First edge element extraction unit, second edge element extraction unit, third edge element extraction unit, and fourth edge element extraction unit 1.61, 62, and 6
3 and 64 represent all connecting lines obtained by the first edge element extraction unit 51, the second edge element extraction unit 52, the third edge element extraction unit 53, and the fourth edge element extraction unit 54, respectively. A first side slope candidate detection unit, a second side slope candidate detection unit, a third side slope candidate detection unit, and a fourth side slope candidate detection unit, which perform linear approximation and detect the slope of the straight line; 7 is a first side slope candidate detection unit; The goodness of fit in each approximation for each straight line obtained by the side slope candidate detection unit 61, the second side slope candidate detection unit 62, the third side slope candidate detection unit 63, and the fourth side slope candidate detection unit 64. an inclination detection unit that calculates the inclination of the rectangular object from the inclination candidates of the rectangular object by assigning a weight according to the inclination of the rectangular object and rotating the first two at right angles to obtain an average value;
81, 82, 83, and 84 are the first side element extraction unit 51, the second side element extraction unit 52, and the third side element extraction unit 5, respectively.
The schools of fish constituting the respective connection lines obtained by the third and fourth side element extraction sections 54 are drawn by straight lines perpendicular to the slope obtained by the slope detection section 7 for the first two, and slopes for the latter two. By scanning a straight line parallel to the slope obtained by the detection unit 7 and searching for the position of the straight line when the number of points whose distance from the straight line is smaller than a predetermined value is maximum, the left side of the rectangular object is searched. , a first side detecting section, a second side detecting section, a third side detecting section, and a fourth side detecting section that detect the right side, upper side, and lower side contours, that is, sides.

In this way, the edge detection section 21 and the edge detection section 22
In this step, image information with a two-dimensional structure is converted into a one-dimensional structure, and therefore, all subsequent processing can be performed at high speed by one-dimensional processing.

C) Effect of the Invention 8- As explained above, according to the present invention, it is possible to obtain a contour extraction device for a rectangular object with improved processing speed.

[Brief explanation of the drawing]

The figure shows a system block diagram of one embodiment of the present invention, 21
・22 is an edge detection unit, 31 work 32, 33r34 are first, second, third, and fourth connection line generation units, 41 and 4
2, 43, and 44 are the first, second, third, and fourth edge candidate generators, respectively, and 51, 52, 53, and 54 are the first and fourth edge candidate generators, respectively.
Second, third, and fourth edge element extraction parts, 61, 62-63,
64 are the first φ second, third, and fourth edge shaving candidate detection units, 7 is the inclination detection unit, and 81, 82, 83, and 84 are the first, second, third, and fourth sides, respectively. This is the detection part.

Claims (1)

[Claims] Using a one-dimensional arithmetic filter, a digital image of a rectangular object is scanned in a rask, and points at which the rate of change in density is at a maximum are extracted one by one for each scanning line. An edge detection unit that extracts fish schools obtained one by one as one-dimensional array data, and connects points whose adjacent distances are equal to or less than a predetermined value among each fish school obtained by the edge detection unit in the order of scanning lines. a connecting line generation unit that obtains a connecting line by extracting the longest connecting line from among the connecting lines obtained by the connecting line generating unit, and obtaining side candidates of the rectangular object by linearly approximating the longest connecting line. an edge candidate generation unit; a side slope candidate detection unit that linearly approximates all the connection lines obtained by the connection line generation unit and the edge element extraction unit and detects the slope of the straight line; a slope detection unit that calculates the slope of the rectangular object from slope candidates for each side of the rectangular object obtained by the side slope candidate detection unit; and a slope of the rectangular object obtained by the slope detection unit and the side element extraction unit. and a side detection unit that detects the sides of the rectangular object from all the obtained connecting lines, and detects the contour of the rectangular object based on the slope of the rectangular object obtained after calculating slope candidates for each side of the rectangular object. A contour extraction device characterized by extracting.