JPH09190536A - Automatic drawing input device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To recognize a graphic by surely separating the vector of noise and a character part from a graphic vector. SOLUTION: One comparison source polyline is taken-out from a table where the polylines are registered (S1 and S2), a contour within distance p1 from the polyline is registered in the second table (S3), whether or not line width, length and angle difference with a comparison source polyline are within a setting range is adopted as a judgement condition (S7-S10) and the polyline which clears the condition is registered as a recognition object (S11).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a drawing automatic input device, and more particularly to a process for recognizing a figure by converting a contour vector of the figure into a polyline.

[0002]

2. Description of the Related Art A drawing automatic input device has a processing procedure shown in FIG. The input topographic map or mechanical / construction design drawing is subjected to an input process of converting it into a binary image by an image scanner or the like (step S1), and an image editing process using the binary image is performed (step S2) to obtain each binary value. Extraction of the contour vector group as a connection of consecutive pixels to the image,
A contouring process for compressing the information is performed (step S
3). Further, core line vector processing for extracting the core line vector group of the graphic line segment from the closed contour vector group is performed. The core vector may be generated at the center by searching for a pair vector from the contour vector.

Contour editing for grouping isolated elements with closed vector groups such as polylines and polygons, and separating drawing elements for classifying character groups and figure candidates (symbol candidates and other line segment candidates) for this group. A process (step S4) is performed.

For each candidate, a recognition process (step S5) is performed, and a drawing such as CAD is performed after an editing process (step S7) including a drawing in which a drawing process (step S6) that is a correction / edit by an operator is performed. Output processing is performed as data (step S8).

[0005]

In the conventional apparatus, when a contour vector is obtained from an image and a contour vector group is converted into a polyline to recognize a figure or a character, all the contour vectors are to be recognized. The image of the noise portion or the character portion due to the dirt may be converted into a figure such as a polyline, and a meaningless contour vector may be recognized as a figure.

In this case, the operator has to perform edit / correction work such as deletion on a meaningless figure, which deteriorates input work efficiency.

An object of the present invention is to provide an automatic drawing input device for recognizing a figure by surely separating a vector of noise or a character portion from a figure vector.

[0008]

According to the present invention, a drawing is read as a binarized image, a contour vector of the binarized image is extracted, and the contour vector is converted into a polyline to obtain a graphic image.
An automatic drawing input device for recognizing characters is provided with means for determining whether or not the polyline is a recognition target figure based on a length and a line width of the polyline, an angle and a distance with a polyline existing in the vicinity of the polyline. And

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a flowchart of a recognition process showing an embodiment of the present invention. In the present embodiment, after converting the contour vector into a polyline, it is determined whether the polyline is a part of the figure based on the length of the polyline, the angle with another polyline existing in the vicinity of the polyline, the distance, and the like. judge.

When recognizing a figure, a large number of polyline data converted from contour vectors are registered in a first table (polyline table). Also, specify the following parameters.

(A) Range (distance) for searching contours p1 (b) Maximum value of line width w1 of recognized polyline (c) Maximum value of length of recognized polyline l1 (d) Recognized as comparison source polyline After registration and designation of the polyline angle difference a1 or more, when the first table in which the polyline data is registered is not empty (S1), one polyline data is taken out from the table and used as a comparison source polyline (S2). . A contour within a distance p1 from this comparison source polyline is searched for and registered in the second table (contour table) (S3).

When this second table is not empty (S
4) Extract one contour from the second table (S
5) Then, the polyline of the contour is recognized as a polyline candidate (S6). When the following conditions are satisfied for this polyline candidate, the polyline is a recognition target.

(S7) The line width of the recognized polyline candidate is smaller than w1.

(S8) The length of the recognized polyline candidate is longer than l1.

(S9) The angle difference between the comparison source polyline and the polyline candidate is smaller than a1.

(S10) The angle difference between the comparison source polyline and the polyline candidate is smaller than (90 ° -a1).

When the above conditions are cleared, it is registered in the first table as a polyline of the figure to be recognized (S1).
1). On the contrary, when the condition is not cleared even under one condition, the polyline is returned to the contour because it is not the polyline of the recognition target figure.

FIG. 2 shows an example of a recognition result based on this embodiment. A contour within a distance p1 with respect to the comparison source polyline is searched, and a polyline candidate a,
Whether or not the recognition target figure is a polyline is determined depending on whether or not the above conditions are satisfied for b and c.

FIG. 3 shows that each of the polylines c, a, and b recognized as the comparison source polyline in FIG. 2 is within the range of the angular difference a1.

[0020]

As described above, according to the present invention, in a drawing automatic input device for converting a contour vector into a polyline and recognizing figures / characters, the length and width of the polyline and the presence of the polyline are present near the polyline. Since it is determined whether or not it is the polyline of the recognition target figure based on the angle and the distance from the polyline, it is possible to reduce erroneous recognition when recognizing drawings that are composed of many lines that are close to horizontal or vertical such as a residential map. However, it is effective in reducing manual correction by the operator.

[Brief description of the drawings]

FIG. 1 is a flowchart of a recognition process showing an embodiment of the present invention.

FIG. 2 is an example of a recognition result in the embodiment.

FIG. 3 is a comparative example of recognition results in the embodiment.

FIG. 4 is a flow of processing of the drawing automatic input device.

Claims (1)

[Claims] 1. A drawing is read as a binarized image,
In the automatic drawing input device for extracting the contour vector of the binarized image and converting the contour vector into a polyline to recognize figures / characters, the length and line width of the polyline, and the angle with the polyline existing in the vicinity of the polyline And a drawing automatic input device comprising means for determining whether or not the figure is a polyline of a recognition target figure from the distance.