基于类内和类间交叉熵的图像分割快速算法
Fast Algorithm for Image Segmentation Based on Cross-entropy Between and In Two Classes
提出了基于类内交叉熵和类间交叉熵的灰度图像分割办法，同时针对抗噪性能进行了二维交叉熵概念的推广，并针对二维直方图分割后不同的区域分别给出了递增和递减两种快速迭代算法，实验结果表明，二维快速递推交叉熵方法能有效的达到分割效果，同时对于含噪声以及目标背景相差较大的图像能更好的达到分割效果。
method based on cross-entropy between and in two classes for gray-level image segmentation is proposed, then extend the concept of cross-entropy to the 2-dimension for the antinoise ability. Besides two fast iterative algorithms based on the different area of the 2-dimension histogram are proposed respectively, one is increased by degree, the other is decreased. The results of the experiment show that the 2-dimension fast iterative algorithm can get a effective segmentation, furthermore it's more available for the case that there is noise or a big discrepancy between the object and background in the image.
丁力、付明明
计算技术、计算机技术电子技术应用
图像分割类内交叉熵类间交叉熵递推算法
image segmentationcross-entropy in two classescross-entropy between two classesiterative algorithm
丁力,付明明.基于类内和类间交叉熵的图像分割快速算法[EB/OL].(2011-01-05)[2025-08-11].http://www.paper.edu.cn/releasepaper/content/201101-195.点此复制
评论