Circular object detection method based on digital image

Abstract

The invention discloses a circular object detection method based on a digital image, which is characterized by comprising the following steps: (1) getting the edge of an image; (2) segmenting the edge into a connected one-way curve; (3) segmenting the one-way curve into a secondary curve; (4) determining an arc; and (5) carrying out circle fitting. Through the method, multiple circular objects can be detected all at a time, and partially blocked circular objects can be detected. Detection is quick and highly precise. The method can be widely used in fields such as industrial circular work piece detection and robot-based circular object recognition, and is of high application value.

Description

technical field [0001] The invention relates to a workpiece detection method, in particular to a digital image circular workpiece detection method. Background technique [0002] Circular object detection is a major focus of industrial inspection and robotics. In industrial production, many workpieces are round or spherical, such as bearings, balls, bottle caps, wheels, etc. The detection and positioning of these workpieces is very important. of. The classic circle detection methods include Hough circle detection and least squares fitting algorithm based on edge tracking. [0003] The basic idea of ​​the Hough circle detection algorithm is to transform a point in the image space into a curve or a surface in the parameter space, and the points with the same parameter characteristics intersect in the parameter space after transformation, and complete by judging the degree of accumulation at the intersection. Detection of characteristic curves. It is an efficient method for d...

AI Technical Summary

Problems solved by technology

But there are also obvious shortcomings: (1) the amount of calculation is large, and the required calculation time is long

(2) It is necessary to pre-store the parameter space unit into the storage unit, which consumes a large amount of storage space

(3) Due to the discretization of image space and parameter space, the detection accuracy of Hough transform is not high

(4) The Hough transform uses a 1-value accumulation scheme to vote equally for all feature points, and cannot distinguish between noise points and feature points that make up the target

But there are also shortcomings: (1) Sensitive to noise, because it is a fitting algorithm, it will naturally try to make the fitting curve close to all the measured data, and when there is noise data, the fitting error will be large

When the measured object contains the edge data of the circle, the fitting accuracy is high and the speed is fast, but when there is background noise, the fitting effect is very unsatisfactory

(2) The least square method fits all points, and only one circle can be detected at a time

Images