Abstract: In digital image processing，the convex and concave shape of objects’ edge，which contain some properties of those objects，can become substantial parameters to describe their shape and some conditions to split some objects which touch together.Some descriptions of the convex and concave shape of edges are described.But those descriptions are not made distinctly，and the computation of the shape excessively depends on pixel unit which is the minimum one in images.This paper firstly makes a definition of the concave point of edge，then according to edge curve simulating methods a new method is given which can exactly anchor the concave point of edge and depends lightly on the pixel unit.In the end of the paper，an experiment is made to explain the application of the new method in the microorganism cell image processing.

摘要： 在数字图像处理的应用中，物体边界凹凸信息通常代表着物体的某些性质，它既可以作为描述物体形状的一个特征参数，也可以作为一些重叠物体分离的依据。人们采用多种方法对边界凹凸变化进行描述，但这些方法对凹点没有严格定义，在计算精度完全依赖像素单位，其计算结果误差较大。首先给出边界凹点的定义，然后基于边界拟合设计一种新的边界凹点定位算法，新算法减少了对数字图像像素单位的依赖。最后通过实验说明新算法在微生物细胞图像分割处理中的应用。