Research On Milk Somatic Cell Image Processing And Counting Method

Somatic Cell Counting (SCC) is an important index of milk quality measurement and bovine mastitis diagnosis. This paper mainly concentrates on the research of pre-processing, segmentation, post-processing and milk somatic cell images counting. And an image processing and SCC software system based on image processing is developed in this paper. The main work is listed as follows.Firstly the weighted mean approach is used to transform color image into gray image, and gauss filter is applied to remove the image noise. An image segmentation approach based on threshold is adopted and a segmentation approach using Otsu method combined with seed filling algorithm is proposed based on character of somatic cell image and the existed cell segmentation approaches. An improved segmentation approach based on image fusion method is adopted to fuse the edge information image and the segmented image, which may obtain segmented cells with smooth and circular edge information. Because of some factors affect accurate counting, the morphology method is applied to complete post-processing and the binary image with the single cell as a connect region is obtained finally.The main goal of the research is to develop a fast and accurate SCC strategy. After analyzing and comparing the existed image marking and counting methods,"Fire Spread"is adopted to SCC. As is shown in the experimental results, it is an appropriate somatic cells counting method with fast and reasonable performance.Finally, SCC software is realized under C++ Builder and a simple somatic cell database is built.