摘要
对模糊C均值算法进行了改进,采用更适合遥感图像的Mahalanobis距离代替欧氏距离,并在聚类中加入了先验信息。在聚类过程中,未标签的样本通过与已标签的样本进行相似性比较来提高算法的准确性。实验表明,改进的算法能有效提高算法准确度。
This article improves the fuzzy C-means method. The improved method adopts Mahalanobis distance which is more likely close to remote-sensing scatter map instead of Euclidean distance. And it adds apriori information into the patterns to change the method as a semi-supervised clustering. In the clustering process, the unlabelled patterns compare similarities with the labeled patterns, then the accuracy of the algorithm can be increased. This method is tested with an experiment, the results shows that the improved method can obviously increase the accuracy of the algorithm efficiently.
出处
《计算机应用研究》
CSCD
北大核心
2006年第7期252-253,260,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60273066)
湖南省教育厅高等学校科研项目资助
