自适应K-means图像分割方法

本文提出了一种自适应K-means图像分割方法,该方法通过用Otsu方法来初步选取阈值作为初始K-means聚类中心,然后经过K-means聚类方法对图像进行分割,最后根据K-means获得的聚类中心与Otsu所确定的阈值进行平均计算,作为图像分割的阈值.实验结果表明:该方法自动快速形成的分割阈值较为合理,对图像能达到更好的分割效果,可以进一步提高后续图像匹配与定位的准确性.

Self-Adaptive K-means on the Method of Image Segmentation

This paper proposes an adaptive K-means about image segmentation method,the method uses Otsu to preliminarily select threshold as the initial K-meansclustering centers,and then the image is segmentated through the K-means clustering method, finally, according to the obtained cluster centers and threshold by Otsu, the average is determined as the threshold of the image segmentation.The experimental results show that:the segmentation threshold is reasonable and the image can achieve better segmentation results