Font Size: a A A

The Research On Inspection Of Gear Quality Based On Computer Vision

Posted on:2008-10-13Degree:MasterType:Thesis
Country:ChinaCandidate:C C RenFull Text:PDF
GTID:2178360212999293Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With rapid development of computer vision technology and photo-electricity technology, the visual inspection technology is quickly developed and widely used. The visual inspection technology as the non-contact measurement method has attracted more and more people's attention. Gear manufacture plays a very important role in modern manufacture. Gear's precision has a great influence on mechanical product's performance. The application of computer visual inspection technology in gear quality inspection can improve gear's precision, raise enterprises'productivity and reduce the costs of production. It has a great significance to improve enterprise's development and market competition.The application of computer vision technology in gear quality inspection is researched and discussed intensively in paper. A system of gear quality inspection based on computer vision is preliminarily developed. Gear's image is processed by system and several diversified algorithms of edge detection are compared, LoG algorithm is chosen in paper because of its good results. The system adopts two methods to inspect the quality of gear product. On the one hand, through analyzing IGES file format, gear's edge image is changed into IGES graphic file with choosing point as entity. IGES graphic not only can be opened but also can be compared with design image. through mechanical design software On the other hand, image registration based on mutual information is applied, the gear edge image and design image are geometrically aligned, the method of image fusion is introduced after that, the result of image registration can be observed in fused image, and whether gear is qualified or not can be decided by rate of registered point in fused image.The system is implemented on C#.net and Matlab platform, the correlative experiments are done, the experimental results which are analyzed in paper prove that the method proposed is feasible. The method presented in this paper can be simply operated and easily observed, so it is a new kind of non-contact measurement technology which has bright development and application future. At the same time, this method has good expansibility, we may apply this method into the quality detection of other mechanical products.
Keywords/Search Tags:Visual Inspection, Edge Detection, IGES, Image Registration, Image Fusion
PDF Full Text Request
Related items