A Estimating Separate Line Algorithm for Overlapping Gastric Adenocarcinoma Cell Image
一种重叠胃腺癌细胞分离线估计算法

Computer-aided medical image processing and analysis is a important application in the field of image processing.The introduction of computeraided diagnosis system of cancerous tissues is highly helpful to clinical diagnosis and medical studies,especially in the case of lacking specialists.But the technology is difficult to put into practice.For example there are many overlapping cells in the gastric adenocarcinoma image because of the facture of the micrograph of gastric adenocarcinoma.If the cells are not separated accurately,the following research will not go on well.It is important to draw the separateline and gain the borderline of each cell.So the paper presents a algorithm which integrated mathematical morphology with an improved B-spline active contour model(B-Snake model) in order to segment the overlapping parts in the gastric adenocarcinoma image.The algorithm adopted an improved nonlinear distance image to force B-spline curve close to the objective boundary automatically.The experimental result of shows that our algorithm can construct the separate-line exactly and draw the original figures of cells effectively.