The Inspection System Research Of Small Modulus Plastic Gear Based On Computer Vision

The city of Ningbo of Zhejiang provice is one of plastic gear production bases in our country. There are hundreds of plastic gear corporations in Ningbo, which own about thirty percent of plastic gear export quantum. So plastic gear has become one of important productions that make profit in foreign currency in Ningbo. In recent years, many producing enterprises have introduced automation line for the production of plastic gear by technical innovations. But the quality inspect of plastic gear have also employed manual workers by using inspecting tools. The result is not only much workload and inefficiency ,but also low reliability and high omission factor ,which is not assorting with the automatic production line. So the plastic gear production corporations are in dire need of the automatic inspection line.As a new non-contact inspection technology, Computer vision inspection is provided huge space by modern industry. Computer vision inspection has the high precision of measurement and the high stabilization of inspection result. It can realize intelligence, network, flexibility , fast and low cost inspection. It also has the capabilities of on-line inspection, real-time analysis and real-time control.The non-contact inspection by computer vision technology can avoid the possible damage of production surface as the contact inspection do, and conquer the subjective defect by manual workers with eyeballing. So computer vision contributes to higher productivity, product quality, product precision and so on.The research and application of international computer vision system has begun from 80 years last century. Computer vision inspection is booming. New conception, theory, technology come forth continuously, which have been applied for vision inspection and automatic recognition of all kinds of products.In our country, industry vision system is still in a time of conception leading. Although the theory and experience of computer vision inspection have gotten some results, it has just started and dropped behind some countries. The lead corporations among every industry have begun to paid attention to automatic vision measurement, after they solved the problems of automatic production.As the detection of metallic gear is a important link, some colleges and scientific research organizations have gradually developed the theory researches of the defect and parameter detections based on computer vision, also have achieved someprogresses and applicationes, which provide good foundation for the quality detection of small modulos plastic gear. Because the detection indices of plastic gear are different from metallic gear's, so it is necessary to profoundly research the computer vision detection theories of plastic gear, which are well applied in the practical detection process.Combining the project of industrial science and technology of Ningbo city(2005B100014)《The inspection system research of surface defect of micro part based on CCD image recognition technology》and the cooperative project of Ningbo academy-corporation《On-line quality inspection of small modulus plastic gear》, the paper profoundly studies the problem of image preprocessing and image segmentation of small modulus plastic gear, the dimension of inner circular orifice and the vision inspection technology of serrate defect of big and small gears.Based on the structure of small modulus plastic gear and the demand of the quality inspections, the paper discusses the hardware's constitute and the principle of selectness of computer vision inspection, especially the design and selectness of lamp- source system,and selects adapted industry digital vidicon , camera lens and lamp- source and so on.The paper selects the small aberration telecentric camera lens, at the same time uses standard specimen and sub-pixel arithmetic to Calibrate system. So it makes the system Calibration fast and precise.The paper particularly analyses the kinds of noise of digital image, and presents some familiar filters used in image preprocessing, and points out the limitations of all filters above by analyzing the characteristic of small modulus plastic gear's digital image.The paper detailedly edge-preserving filter, which is well used for the image preprocessing of small modulus plastic gear.The image segmentation of big and small gears is one of difficulties in the paper. The paper presents threshold iteration, maximum classes variance and probability relaxation and so on. In the process of researching the image segmentation algorithm, the paper applies many algorithms and improved on some of them. Although the result of image segmentation is satisfied, the running speed of the algorithm can't accord with the request of real-time inspection because of the complexity.Considering the characteristic of small modulus plastic gear, the paper presents the incision synthetic method of threshold iteration with maximum classes variance,which solves the multithreshold problem with synthetic single-thresholds method. The synthetic algorithm runs faster according with the real-time inspection, also solves the difficulty of big and small gears'image segmentation.Because of the dust on transmission line, lamp- source, camera lens, background image appears some motley point groups that are not eliminated easily in the process of image preprocessing. The motley point groups still exist in the image after image segmentation, and affects the following image inspections. The paper presents the removing algorithm of motley point groups, so strengthens anti-jamming of the computer vision inspection system.The paper further studies the pixel connectivity of digital image, and discusses the predigestion algorithm of contour extraction and optimation. It gets the satisfied optimization contour to achieve the correct dentiform inspection's data.The paper achieves the coarse positioning of small modulus plastic gear's inner circular orifice by the method of determining circle with three points. It provides primary data for sub-pixel positioning. The paper discusses gray moment- preserving edge positioning method and least-square circle method, which provide theory for researching appropriate sub-pixel positioning method. The paper applies the synthetic algorithm of gray moment-preserving edge positioning with least-square circle,which achieves sub-pixel positioning of the inner circular orifice, and also applies 3σiteration principle to improve the precision of sub-pixel positioning. The paper presents dummy circle scan method to achieve dentiform inspections of all kinds of big and small gears.The computer vision algorithms as raised in this paper are well applied in the practical detection system of small modulos plastic gear.