Image acquisition: The fastener is illuminated by the light source system, and the fastener is photographed by an industrial camera to obtain its image information. The choice and arrangement of light source is very important, it will affect the quality of image and the accuracy of feature extraction.
Image processing: The captured image is transmitted to the image processing system, the system uses a variety of image processing algorithms, such as grayscale, filtering, binarization, edge detection, feature extraction, etc., to analyze and process the image to extract the key feature information of the fastener.
Feature comparison and judgment: the extracted features are compared with the pre-set standard template or parameters, and the technology of pattern recognition and deep learning is used to determine whether the fasteners have defects and whether the size meets the requirements, so as to obtain the detection results.