Wear particle image segmentation using a two-stage strategy
Heng-guangGuo1, Jun Qu2
COMPUTER MODELLING&NEW TECHNOLOGIES 2014 18(12B) 616-622
1Graduate Students’ Brigade, Naval Aeronautical Engineering Institute, Yantai, 26400,China
2Departmentof Airborne Vehicle Engineering, Naval Aeronautical Engineering Institute, Yantai, 264001, China
Wear particle image analysis is an effective and reliable method for equipment condition monitoring and fault diagnosis. Segmenting wear particle from image is an important but challenging problem. In this paper, a two-stage wear particle image segmentation strategy is presented, which consists of a rough segmentation stage and a fine segmentation stage. In the first stage, a wear particle image is divided into blocks, and clustering method is used to group blocks. This stage aims to get the rough boundary of the wear particle. In the second stage, color gradient is introduced into GVF snake to establish colorGVF snake model, and rough boundary from the first stage is used as initial contour. This stage tries to extract the accurate boundary of the wear particle. Experimental results shows that the method proposed in this paper offers an accurate, minimally interactive, and efficient scheme for wear particle image segmentation, and increases the quality of wear particle image segmentation in compare with some state-of-art segmentation methods.