中文摘要：

在研究K均值聚类算法的基础上,采用小波变换辅助K均值算法对遥感影像进行分类,以此提高遥感影像的分类精度.以云南省玉溪市抚仙湖附近地区作为研究区,结合研究区的具体情况,根据查维茨最佳指数因子法OIF计算得到遥感影像的最佳波段组合,并通过对各类地物的样本图像和遥感影像进行二维小波分解,得出样本特征向量;然后利用K均值算法结合样本特征向量对遥感影像进行分类,得到分类结果并进行精度验证.再与单纯采用K均值算法的分类结果进行对比分析,结果表明：其总体精度和Kappa系数分别达到83.74%、0.7753,比单纯采用K-means算法分别高出14.26%、0.1697,尤其是林地、裸地和农田的分类精度得到了显著提高.

英文摘要：

On the basis of studying the K-means clustering algorithm,combine wavelet transform is combined with Kmeans algorithm for remote sensing image classification to improve the classification accuracy of remote sensing image.Fuxian Lake area in Yuxi city of Yunnan Province is taken as a study area,combined with the specific circumstances of the area,the optimal bands combination of remote sensing image is obtained according to the OIF calculation.Through the twodimensional wavelet decomposition of various terrain samples and remote sensing image,the sample feature vector is obtained.Using K-means algorithm with the sample feature vector for classifying the remote sensing image,the result of image classification is got and the accuracy is verified.Comparing with the classification result using K-means algorithm simply,the results show that its overall accuracy and Kappa coefficient are 83.74% and 0.7753 respectively,increasing by14.26%,0.1697.Especially the classification accuracy of forest land,bare land and farmland is greatly improved.