二维Renyi灰度熵和模糊聚类在图像分割中的应用

Application of 2D Renyi Gray Entropy and Fuzzy Clustering in Image Segmentation

针对传统二维Renyi熵阈值算法计算复杂度高的问题,提出一种改进的二维Renyi图像分割算法。该算法利用两个一维Renyi熵算法的阈值来代替传统二维Renyi法的阈值,使得计算复杂度大大降低;同时为了改善图像分割效果,综合使用了FCM算法。实验结果表明,改进的算法充分发挥了二者的优势,不仅提高了运算速度,且分割效果更好。

Because of the high calculating complexity of classical two-dimensional Renyi entropy thresholding, an improved algorithm was proposed in the paper. Instead of calculating the traditional 2D Renyi threshold, it reduced the complexity by computing two 1D Renyi threshold. In order to improve the global segmentation performance, FCM（ Fuzzy C-means Clustering） was adopted in the algorithm. Experimental results showed that this improved algorithm gave full play to the advantages of both, not only in the computing speed, but also in the quality.