Watershed segmentation of SAM cell images

• 1、College of Electrical and Information Engineering, Hunan University, Changsha 410082

Abstract：Using image processing technology to identify and track cells in cell image, is one of the important research topic in the field of computer vision. Medicine or biology researchers can obtain rich cell dynamic informationthrough the automated or semi-automated cell tracking methods, and cell recognition is the basis of cell tracking. In multi-cell detection aspects, we firstly studied automated segmentation algorithm for cell microscopic image, and analyzed the characteristics and application of level set method and watershed transform method. In order to improve the segmentation accuracy using watershed method, we designed a mixed filter which contains wavelet denoising and average filter before final segment. The segmentation algorithm uses the advantages of combining spatial domain filtering and frequency filtering to denoise the source image, solve the segmentation error of over-segmentation and under-segmentation, and ensure that the correct target information is highlighted.

Keywords： Image processing Over-segmentation Under-segmentation Watershed algorithm