基于分水岭算法的双向凝胶电泳图像分割

2D-gel Electrophoresis Image Segmentation Algorithm Based on Watershed

采用分水岭算法对电泳凝胶图像进行分割,针对其存在的过分割问题以及图像本身的特点,提出了根据拓扑曲率对过分割区域进行合并,以得到有意义的分割结果。用本算法对电泳图像进行了分割实验,结果表明,用本算法适用于电泳图像,并且效果良好。

In this paper, an algorithm based on watershed transformation has been applied, which is adapted to 2D-gel electrophoresis image segmentation. According to the over-segmentation of watershed algorithm and the characters of 2D-gel electrophoresis image, the means of curvature on topology has been presented, which was used to unify the over-segmented regions into some significant regions. Finally, the means was verified through some experiments and have been given good results.