Circle detection method based on histogram peak value search

Abstract

The invention discloses a circle detection method based on histogram peak value search. The method includes the following steps of firstly, conducting edge direction on an original image to obtain a binary image; secondly, horizontally and vertically scanning edge points in all lines and all rows in the binary image to obtain a horizontal and vertical midpoint histogram; thirdly, searching for an overall peak value in the horizontal and vertical midpoint histogram wherein the transverse coordinate corresponding to the overall peak value and the vertical coordinate corresponding to the overall peak value are coordinates of the circle center of a candidate circle; fourthly, obtaining a histogram of distances between all the edge points and the circle center of the candidate circle and normalizing the histogram; fifthly, judging whether the candidate circle is true or false according to the local peak value in the normalized distance histogram; sixthly, deleting the edge points located on the candidate circle, renewing the horizontal and vertical midpoint histogram, returning to the third step, and repeatedly executing the steps till the number of detected circles meets the requirement. By means of the method, the geometrical feature that a circle is centrosymmetric is fully used, the circular target can be rapidly and accurately extracted based on histogram peak value search, and compared with a common method, the efficiency is improved by a magnitude order.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a circle detection method based on histogram peak search. Background technique [0002] Detecting circular objects in digital images is a classic problem in computer vision and pattern recognition, and is widely used in automatic detection, digital image processing, medical image analysis and other fields. Hough transform is a method commonly used in image feature detection and recognition. It has the advantage of being insensitive to noise and can effectively filter out the influence of noise to improve the accuracy of detection results. However, when it is used for circular target detection, the parameter space exceeds two dimensions, the amount of calculation and storage space are relatively large, and the efficiency is low in practical applications. In recent years, in-depth and extensive research on circle detection methods has been carried out at home and ...

AI Technical Summary

Problems solved by technology

In recent years, in-depth and extensive research on circle detection methods has been carried out at home and abroad, and many achievements have been made: Random Hough Transform (RHT) randomly selects points, when the content in the image is more complex, it will cause a large number of invalid sampling and accumulation, affecting the detection Efficiency; its improved algorithm uses methods such as density checking, cluster analysis, and sampling optimization to effectively reduce invalid accumulation, but increases the complexity of the algorithm; some other algorithms have higher detection accuracy, but need to calculate derivatives, gradients, connectivity, etc. Mathematical operations, the specific implementation is more complicated

Images