文献类型：期刊文献

英文题名：A mechanical parts inspection model based on image classification technology

作者：Du, Xin[1]

第一作者：杜鑫

通讯作者：Du, Xin

机构：[1] Department of Mechanical and Electrical Engineering, Xinxiang University, Henan, China

第一机构：新乡学院机电工程学院

年份：2016

卷号：14

期号：3

起止页码：14-21

外文期刊名：Academic Journal of Manufacturing Engineering

收录：EI(收录号：20164903099242);Scopus(收录号：2-s2.0-85000398878)

语种：英文

外文关键词：Inspection

摘要：In traditionally, the mechanical parts inspection is through manual operation, which highly requires operators a good mentality, working experience, physical condition, education, professional ethics and skill. Manual inspection deeply depends on the individual's experience, and is not replicable, coupled with increasingly employee turnover, which leads to instability of the products inspection. As such, the computer image classification technology soon becomes the best option for inspecting the quality of products with its advantages of fast, flexible, intelligent and not subject to physical limitations. In this paper, process quality check system is designed based on the image classification technology. The technology that is adopted in the experiment includes the inspection of interest point, the description of interest points and the visual dictionary generation to check the effectiveness of the system. As the result of experiment, the use of computer vision for the qualitative inspection of mechanical parts can improve the efficiency and accuracy of inspection to a certain extent.