SU754453A1 - Method of discriminating image contours - Google Patents

Description

The invention relates to automation and computing and can be used to minimize visual information when solving recognition problems.

Known methods for the selection of the contours of the image, which consists in the selection of signals of positive and negative images, the redistribution of negative image signals, their subsequent summation.

The disadvantage of these methods is low reliability in the case when it is required to distinguish the image signal of low-contour contours from the background signal and their unsuitability for distinguishing the contours of three-dimensional objects [1].

The closest to the invention to the technical essence is a method for the selection of the contours of the image, which consists in the discrete conversion of the optical image into electrical signals and the formation of difference signals from adjacent parts of the image [2].

However, this method is impossible to distinguish the contours of images of three-dimensional objects.

2

The aim of the invention is to expand the scope by highlighting the contours of images of three-dimensional objects.

This goal is achieved by the fact that ₅ in the proposed method for extracting image contours, including operations of discrete conversion of an optical image into electrical signals and generating difference signals from adjacent image areas, additionally introduced defocusing operations of the image and determining the presence of contours from the maxima of difference signals adjacent

sections of the image.

FIG. 1 shows a block diagram of a device for implementing the proposed method ¹⁵ ; in fig. 2a is a graph of the change in focal length of the optical system of the dynamic defocusing unit; on . FIG. 26 - dependence of the voltage at the output of the ί-th comparison element with dynamic defocusing; in fig. 2c is the dependence of the voltage at the output of the ίth pulse shaping unit on the amplitude of the voltage at the output of the ϊth comparison element.

754453

3

The device contains (Fig. 1) dynamic defocusing unit 1, information reading unit (matrix of photoelectric converters 2, comparison elements 3, pulse shaping units 4 and decision block 5.

The device works as follows.

The image of the object is projected on the matrix of photoelectric converters of the information reading unit 2.

Dynamic defocus unit 1,

carries out periodic, with a period T, the defocusing of the image contours on the matrix plane of photoelectric converters of the information reading unit 2. The graph of the change of the focal distance G of the dynamic defocusing block 1 ^{1 is} shown in FIG. 2a There is a dependency

± ₌ 3__X

η ί ’

, · ²

where a;, is the distance of the v-element of a three-dimensional object to the optical system of the dynamic defocusing unit 1;

ί is the distance from the image plane of maximum sharpness to the optical system of the dynamic defocusing unit 1; P is the focal length of the optical system of the dynamic defocus unit 1, corresponding to the maximum sharpness of the image of the ϊ-th element of a three-dimensional object. ,

At the time ίχ for the period T (Fig. 2a), the elements of a three-dimensional object, the distance from which to the optical system of the dynamic defocusing unit 1 is, have the maximum sharpness of the contours 2 of the image. The outputs of the two adjacent photoelectric converters of the photomatrix of the information reading unit 2 are connected to the inputs of the ί-th element of comparison 3, voltage I (whose output has a maximum at the time point And (Fig. 26) corresponding to the maximum sharpness of the corresponding ί-th image contour lying on the two

four

adjacent photoelectric converters of the photomatrix of the information reading unit 2. The pulse shaping unit 4 at the time point And corresponding to the maximum voltage I ^ at the output of the ί-th comparison element 3, gives a short pulse A (Fig. 2b).

The presence of a dynamic defocusing of a voltage pulse Αζ at the output of the ί-th pulse shaping unit 4 at the time Ц for the period T means the presence on the corresponding photoelectric converters of the photomatrix of the information reading block of the 2-th contour of the image of the element of a three-dimensional object defocusing 1 equals b;,.

Thus, in addition to the 2 signals generated at the output of the reading unit, which carry information about the shape of the contours of the elements of a three-dimensional object, signals that carry information about the location of the image contours in three-dimensional space are also produced at the output of the pulse shaping units 4.

Claims (1)

Claim A method for isolating image contours, which consists in discrete conversion of an optical image into electrical signals and generating difference signals from adjacent image areas, characterized in that, in order to expand the area of application, by isolating image contours of three-dimensional objects, defocus the image and determine the presence of contours by maxima signals of the difference of adjacent parts of the image.