Detecting Technology Of Fracture Splitting Groove Of Car Connecting Rod Based On Computer Vision

The automobile manufacture has become a pillar industry in the globaleconomy. With the rapid rise of the automotive industry production, it has putforward higher requirements for the production and detection technology relatedparts and components products. As one of the key parts of the engine, the connectingrod requires strictly processing technology because of its high accuracy. The fracturesplitting technology is one of them, since the technique was introduced into China, ithas attracted more and more attention and research. However, the detection whetherthe fracture splitting groove is processed and qualified still remains in the eyeballingobservation and testing phase, which is time-consuming, low efficiency and morepersonal errors. In view of this situation, an automatic detection technology based onmachine vision is discussed in this article.Machine vision refers to the use of computers, cameras and other machinesinstead of human eyes automatic measurement, judgment, recognition and otherfunctions of the target feature. Therefore, a hardware scheme of detecting system isdetermined based on the detailed research and analysis on the characteristics of thefracture splitting groove, which consists of CCD camera, computer,1394imageacquisition card, PLC, controllable light and others. After PLC receiving thedetection signal, controllable light is opened. By controlling the CCD cameraexposure, image of the measured target is acquired. Industrial computer realizes theautomatic recognition and detection of fracture splitting groove of automobileconnecting rod by processing and detecting the collected image.In order to completing automatic extraction and recognition quickly andaccurately, the article chooses the most appropriate algorithm based on discussingand simulating filter algorithm, segmentation algorithm, feature extraction and recognition algorithm. First image is filtered by homomorphic filter based onwavelet transform, and then segmented by self-adaptive canny edge operator, finallythe edge is extracted by morphological operations, boundary line is described byFreeman coding method, and characteristics of fracture splitting groove is describedby feature recognition. Based on these theories, the detection software system forfracture splitting of connecting rod is developed by using VC++on Microsoft VisualC++2005Software development platform and OpenCV component which isdedicated to image processing.Finally, we make a large number of detection experiments on automobileconnecting rod using the visual detection system. Through a detailed analysis of thetest process and results, the visual detection system for fracture splitting groove ofconnecting rod can completely detect whether the fracture splitting groove isprocessed and qualified quickly and accurately. By repetitive detection test for a longtime, the system is proved good stability and reliability. So it can meet the needs ofpractical production.