Automatic Choroid Cells Segmentation And Counting Based On Approximate Convexity And Concavity Of Chain Code In Fluorescence Microscopic Image

With the number of the people who suffering from various retinal diseases is increasing today, the study of diagnosis and treatment of the retinal diseases is significant. In the process of restoration of retinal diseases, the idea of the research is injecting the starch material which can contain the retinal cells into the eyes. Because of the complexity and adhesive of the cells, the work of cell image processing is difficult. Dealing with the medical images in computer technology can reduce the subjective interference and ease the burden of doctors, and then improve the work efficiency. In biomedical image study, the segmentation of cell images automatically has a important influence on disease diagnosis, quantitative analysis of cell information, the visualization of cell microstructure and so on.There are two kinds of work in my study: the acquisition of retinal cell fluorescence microscopic image and automatic segmentation and counting in cell image. In the process of restoration of retinal diseases, the idea of the research is injecting the starch material which can contain the retinal cells into the eyes. Then cells grow and proliferate on the damaged parts. The repairation of the damaged tissue will be a great help to the treatment of diseases. By comparing the number of the living cells growing with the material and without the material, we can judge the poisonousness of the material, and prepare for the next research.After the fluorescence microscopic cell image be obtained, the cell image should be analyzed. In this paper, a method based on the Freeman chain automatically for fluorescence microscopy images was proposed. The proposed method consists of four main steps. First, a threshold filter and morphological transform were applied to reduce the noise. And the adhere cells were detected based on the area and shape of the cells. Second, the boundary information was used to generate the Freeman chain codes. Third, the concave points were found based on the relationship between the difference of the chain code and the curvature. Finally, cells segmentation and counting were completed based on the characteristics of the concave points. The proposed method was tested on 100 fluorescence microscopic cell images, and the average true positive rate(TPR) is 98.13% and the average false positive rate(FPR) is 4.47%, respectively. The results show the effective of the method.