Abstract: A segmentation method of overlapping protein spots is proposed for detection of overlapping protein spots in two-dimensional gel electrophoresis images, which is based on the combination of corner detection and polygonal approximation. First, Harris corner detector is used to distinguish the corner points from the overlapping protein spots. Second, the corner points are used to make polygonal approximation of protein spots and the concave corner points are obtained by judging the convexity and concavity of polygonal vertices. Finally, overlapping protein spots are separated by constructing the segmentation lines while selecting the concave points from the concave corner points under the concave points selecting and matching rule. Experimental results show that the method can be used to separate the overlapping protein spots accurately, especially the severe overlapping protein spots.

Key words: overlapping protein spots, corner detection, polygon approximation, concave point , segmentation

摘要： 针对双向凝胶图像中的重叠蛋白质点检测，提出一种角点检测与多边形近似相结合的重叠蛋白质点分割方法。首先利用Harris角点检测法得到重叠蛋白质点的角点，然后利用角点对蛋白质点进行多边形近似并判断多边形顶点的凹凸性获得凹角点，最后通过凹点选取与匹配原则从凹角点中选取凹点并构造分离线完成对重叠蛋白质点的分割。实验结果表明该方法能准确快速地实现重叠蛋白质点的分割，尤其适用于重度重叠的蛋白质点。

关键词: 重叠蛋白质点, 角点检测, 多边形近似, 凹点, 分割