一种基于最大类间方差和区域生长的图像分割法

提出一种基于一维最大类间方差和区域生长的图像分割法。首先用一维最大类间方差法确定最佳分割阈值,再用改进的区域生长法分割得到目标。实验结果表明,该分割算法不仅适用于简单的图像分割问题,而且对于背景复杂、光照不均匀的图像也能取得较好的分割效果。该算法计算量小,实时性和分割精度均有一定优势,在提取目标的同时,不留下任何背景像素,使下一步的目标识别更为简单。

An Image Segmentation Algorithm Based on Maximal Variance Between-Class and Region Growing

A segmentation algorithm based on the technique of one-dimension maximal variance between-class and region growing method is proposed in this paper.Firstly,the method of one-dimension maximal variance between-class is used to obtain the optimal segmenting threshold.Then,the target is segmented from the original image by the improved region growing method.The experiments indicate that the segmentation method presented in this paper is not only fit for the segmentation of the simple image,but also fit for those images with complex background and uneven light.Moreover,this segmentation algorithm has advantages of real time and segmentation precision.With this method,the target can be extracted without any pixel of the background.Therefore,the target recognition in the next step will be simple.