Title :
The new detection system for missing bearing´s rolling elements in double row ball bearing
Author :
Xu Qiaoyu ; Peng Ping
Author_Institution :
Coll. of Mechatron. Eng., Henan Univ. of Sci. & Technol., Luoyang, China
Abstract :
In the assembly process of double row ball bearing, the lack of rolling elements is the most serious deficiency problems. The current equipment based on machine vision technique can only detect single row except double-row one. This paper establishes a new detection method and system to supply a gap. The detection system uses digital image processing technology to finish identification for the lack of rolling elements quickly and accurately. Firstly, multiple concentric annulus are used to locate precise position of the bearing; and then the image is scanned in the direction of roundness- expansion to extract rolling elements area and the number of rolling elements are counted by statistics. Finally, the position of missing rolling elements is located by standard block scanning, and it will used to display marked result after coordinate transformation. Experiments has shown that the system is efficient, strong anti-interference, etc., have a good prospect of industrial application.
Keywords :
ball bearings; computer vision; image segmentation; rolling bearings; bearings; detection system; digital image processing technology; double row ball bearing; machine vision technique; rolling elements; roundness-expansion; Computers; Image edge detection; Low pass filters; Machine vision; Pixel; Reliability; Transforms; bearing; coordinate transformation; detection; rolling elements;
Conference_Titel :
Computer Science and Automation Engineering (CSAE), 2011 IEEE International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-8727-1
DOI :
10.1109/CSAE.2011.5952656