Research on Sub-Pixel Edge Detection Technology of Piston Surface Defect Image
Aiming at the problems of low contrast,small defect area,many kinds of defects and low efficiency of manual detection in piston image,a piston surface defect detection method combining region growing method and sub-pixel edge extraction is pro-posed.The piston image was collected by image processing software,the difference of gray value between defect area and normal area in piston surface image is analyzed.The region growing method is used for image segmentation,and the Canny operator is combined to preliminarily locate the surface defect edge of the piston.By locating the location of the region of interest,sub-pixel level extraction is carried out,and the edge of the region of interest is smoothed.Experimental results show that the proposed algo-rithm is more accurate and smooth than the traditional edge extraction method.
Region Growth MethodSub-Pixel Edge DetectionRegion of InterestImage SegmentationDefect Detection
NETL
NSTL
万方数据
维普
