摘要
针对高分辨率遥感影像建筑物提取时易受噪声影响,且灰度共生矩阵仅通过单波段进行纹理特征提取导致信息不足等缺点,提出了利用最小噪声分离变换降维去噪,从而提高灰度共生矩阵纹理特征提取效果,并对影像进行双边滤波处理,最后,利用支持向量机算法提取融合了特征信息的多光谱影像中的建筑物信息的方法。实验结果表明,改进后的方法相对于传统方法对建筑物识别精度高、效果好。
Due to the shortcomings of high-resolution remote sensing images,which are easily affected by noise when buildings are extracted,and gray scale co-occurrence matrix is used to extract texture features only through a single band,a method is proposed to minimize the noise and Improve the grayscale co-occurrence matrix texture feature extraction effect,and bilateral filtering of the image.Finally,extracting building information in multispectral images that incorporates feature information by using a support vector machine algorithm.The experimental results show that the improved method has higher recognition accuracy and better effect than traditional methods.
作者
朱金山
宋珍珍
纪轩禹
ZHU Jinshan;SONG Zhenzhen;JI Xuanyu(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China;State Key Laboratory of Geographic Information Engineering,Xi′an 710054,China;Key Laboratory of Surveying and Mapping Technology on Island and Reef,Ministry of NaturalResources,Qingdao 266590,China)
出处
《测绘与空间地理信息》
2019年第7期8-10,共3页
Geomatics & Spatial Information Technology
基金
地理信息工程国家重点实验室开放研究基金资助项目(SKLGIE2017-Z-3-3)
2015测绘地理信息公益性行业科研专项(201512034)资助
关键词
MNF变换
SVM
纹理特征
建筑物提取
MNF transform
SVM algorithm
texture
feature building extraction
