ass日本风韵熟妇pics男人扒开女人屁屁桶到爽|扒开胸露出奶头亲吻视频|邻居少妇的诱惑|人人妻在线播放|日日摸夜夜摸狠狠摸婷婷|制服 丝袜 人妻|激情熟妇中文字幕|看黄色欧美特一级|日本av人妻系列|高潮对白av,丰满岳妇乱熟妇之荡,日本丰满熟妇乱又伦,日韩欧美一区二区三区在线

With the rapid development of remote sensing technology, remote sensing image resolution has been greatly improved and the ground targets can be obtained from high resolution remote sensing image. But the traditional visual interpretation has low work efficiency and needs for professional knowledge. Thus using high resolution remote sensing image to construct automatic interpreting model is the key to quickly and efficiently obtain the building of hollow village. Meanwhile, it is important for the hollow village renovation and research. Based on this, a novel automatic building interpretation model of hollow village high resolution images based on bagofwords (BoW) was proposed. Firstly, several existing feature extraction methods were compared, and then based on the BoW and support vector machines (SVM) the automatic interpretation model for the building was constructed. In order to verify the validity of this method, the high resolution remote sensing image of typical hollow village was selected to construct the building sample library. Finally, the model for building interpretation was experimentally studied based on the sample library. The results showed that the classification accuracy (ACC) of this method can reach 0.86. Therefore, the proposed method can be used for the building automatic interpretation, and it had high practical value to hollow village research and renovation.