AUTOMATICDETECTION OF OIL VOLUME IN CHECKOUT OF OIL PUMPS BASED ON MACHINE VISION

• 作 者: LIUZhao-Dong;LIAOQiang;YANGZheng-Hua;WENRong;
• 机 构: College of Mechanical Engineering,Chongqing University,Chongqing 40044,China
• 出 处: INternational Conference on Product Design and Manufacturing
• 发表日期: 2007年期

　　In order to measure the micro volume of oil,promote the accuracy of oil measuring and the degree of automation in the process of traditional oil pumps correction,an automatic detection system based on machine vision has been established,In this System,CCD camera is treated as sensor and the collected image ofoil level is regarded as the object of the study.After some preprocessing of the image,we improve Canny Margin Extraction to achieve a better effect on margin detection.And then we use least-squares fitting to detect the sub-pixels margin line of the image.Three major characteristics of the system are as follow:first,the improved Canny Edge Detection method with higher precision can achieve size and position detection in sub-pixel level; Second,non-contact measuring of micro volume oil has been achieved by machine vision detection which can detect the parameter that most existing sensor cant inspect,and automatic measurement of the oil volume in oil pumps checkout has been achieved; third,the improved Canny Edge Detection can be used extensively in the precise measurement and accurate position in other projects.