作者: Jun Rong , Genyun Sun , Aizhu Zhang , Hui Huang
DOI: 10.1007/978-3-030-39431-8_24
关键词: Mean shift segmentation 、 Hyperspectral imaging 、 Sparse approximation 、 Segmentation 、 Extraction (military) 、 Pattern recognition 、 Impervious surface 、 Computer science 、 Artificial intelligence
摘要: Impervious surface is an important factor in monitoring urban development and environmental analysis. However, spectral differences structural exist on impervious surfaces, which leads to accurate extraction of surfaces a difficult task. Therefore, this paper proposes superpixel sparse representation based morphological profiles raw data extract the hyperspectral imagery. Specifically, segmentation map image first generated by mean shift segmentation. Then, attribute are extracted stacked with data. Finally, masked onto each resulting classified via representation. Experiments show that method has good performance advantages comparison method. This shows effectiveness proposed