A Computer Vision Technology Applied in the Task of Mechanical Parts Inspection

Abstract

In this paper, based on the computer vision detection technology, the parts detection system is studied and discussed. This system mainly uses manual image segmentation and other methods, which can fully avoid the defects of image segmentation, improve the accuracy of image detection and ensure the detection speed. The system uses CCD or CMOS digital camera to capture the image of parts on the assembly line, which is roughly divided into three parts. First of all, it is necessary to take the template image that fully meets the quality requirements of the parts, and carry out preprocessing such as smooth filtering for the image taken. Second, during the test, the image of the component to be tested is taken on the assembly line, and the position of the detection target image and the standard template image is registered through the image registration algorithm. Finally, the image features of each partition area of the image to be detected are extracted and compared with the features of each partition area of the standard template image, so as to detect the processing quality of the assembly line parts, whether there is a component assembly error, error prompt and alarm, so as to achieve intelligent detection Purpose.