Research And Application On Key Algorithms Of Image Processing In Vision Measurement

With the applications of machine vision technology in inspection field, vision measurement technology has developed rapidly and become one of the important research fields in instrument science. Vision precision measurement technology has been studied deeply abroad, and also has gained more and more attention in domestic. Because of its flexibility, rapidness, non-contact, precision, high automation, vision precision measurement has been widely used worldwide.Image processing technology is an important part of the vision measurement system. Its main task is processing and extracting effective information of images, thus high-precision measurements is carried out. In this paper, the key algorithms of image processing in vision measurement were studied deeply, including: image filtering, image segmentation and edge detection. Based on the knowledge of image noise and filtering, two fuzzy filtering algorithms were proposed to process hybrid noise. The subjective and objective evaluations of filtering quality were presented in subsequent experiments. Through the study of classical image segmentation algorithms and interval-valued fuzzy sets (IVFS), the rule with the largest index of ultrafuzziness of IVFS was used to obtain the best threshold. The approach achieved a relatively desired result in processing reflection sites of the workpiece. For simple shape of workpiece and high contrast in images, system requires high measurement speed. The morphology algorithm was utilized to process the images after segmentation. The edge acquired by this approach can meet the requirement of vision measurement.