A Circle Center Location Method Based on YOLOv5 Algorithm
Accurately obtaining the center of a circular object in an image is a key issue in object recognition and local-ization.At present,the least square fitting circle and HOUGH transform methods are mainly used to locate the center of a circle.However,these methods have limitations in varying degrees,such as weak robustness,high requirements for ambient light,the need to debug parameters in advance to determine the threshold value,and sharp deterioration of the effect un-der complex background.In response to this issue,this paper proposes a comprehensive application of YOLOv5 algorithm,Grabcut algorithm,and grayscale centroid method for image classification,image segmentation,and grayscale centroid al-gorithm to obtain the center of circle coordinates.This method uses YOLOv5 to coarse locate circular targets in the image,then segments circular targets through image segmentation methods.Finally,a weighted grayscale center of gravity algo-rithm is used to accurately locate the center coordinates of the circle,achieving reliable positioning of circular markers.This paper compares the accuracy and stability of our algorithm with existing algorithms through multiple experiments.
center positioningYOLOv5 algorithmimage segmentationgrayscale centroid method
万方数据
维普
