欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
介绍了一种基于Wedgelet（楔波）变换的遥感图像分类算法．该算法将多尺度Wedgelet变换应用于遥感图像区域分割，在此基础上提取各分割区域的Gabor纹理特征实现对遥感图像的分类．为了检验该算法的可行性，将其应用于向海和查干湖遥感图像，并与灰度共生矩阵、高斯马尔科夫随机场等纹理分类算法进行了比较．结果表明，该算法要优于灰度共生矩阵及高斯马尔科夫随机场分类算法，能够得到较高的分类精度和Kappa系数．
英文摘要：
A classification algorithm based on Wedgelet transform for remote sensing images was proposed. In the algorithm, the multi-scale Wedgelet transform was applied to segment the remote sensing image, and then texture features were extracted from the segmented region to perform the classification for the remote sensing image. To illustrate the efficiency of the algorithm, experiments were carried out to the remote sensing images of Xianghai and Chagan lake respectively. Results were compared with the texture classification algorithms of Gray Level Co-occurrence Matrix （GLCM） and Gauss Markov random field （GMRF）. The results show that the algorithm proposed here performs better and it has a higher classification accuracy and Kappa.
同期刊论文项目
同项目期刊论文
The hidden neurons selection of the wavelet networks using support vector machines and ridge regress
期刊信息
位置：