基于OTSU算法的图像分割有效性研究

Research on Image Segmentation Effectiveness Based on OTSU Algorithm

基于阈值分割的二维OTSU算法的基础上,提出了关于二维OTSU算法的快速实现。该算法既利用了图像的像素点灰度信息,同时又利用了像素点的邻域空间相关信息。对于低对比度、低信噪比的目标,二维最大类间方差算法具有较高的分割效果。实验结果表明,二维最大类间方差算法的快速实现方法使图像分割的有效性得到很大提高,极大地扩大了二维OTSU算法的适用范围。

Based on one-dimensional OTSU threshold segmentation algorithm and two-dimensional OTSU algorithm, the fastimplement method of two-dimensional OTSU algorithm is presented.The algorithm uses the image pixel gray level information and thepixel＇ s neighborhood space information.For low contrast and low SNR,the two-dimensional between-cluster variance algorithm has goodsegmentatiom results. The experimental results show that the fast implement method of two-dimensional between-cluster variancealgorithm improves significantly the image segmentation effectiveness and extends the application scope of two-dimensional OTSU algorithm.