Wear Particle Chain Segmentation Based on the Nearest Neighbor Method

Wear particle segmentation is an important step in the analysis and processing of ferrographic images, and it is also a hot topic in the field of ferrographic images. At present, the acquisition of ferrographic images is mostly based on the principle of magnetic field deposition. Wear particles will be chained and accumulated during the deposition process. Therefore, an effective wear particle segmentation method is needed. In this paper, a wear particle segmentation method based on the nearest neighbor algorithm is proposed. The method first decomposes the captured video into images. Then, this method introduces the nearest neighbor algorithm to extract the deposition process of wear particles, uses the distance transformation to form markers, and uses the marker-controlled watershed to solve the segmentation of the wear particle chain.Compared with traditional watershed segmentation algorithm, the problem of over-segmentation and under-segmentation is solved. The experimental results show that the segmentation results of the ferrographic image are accurate and fast, which lays a foundation for the subsequent extraction of the wear particle features.