基于光谱整体形态和局部吸收谷位置的高光谱遥感图像地物识别

Object Identification of Hyperspectral Image Based on the Spectral Overall Shape and Local Absorption-band Positions

为了精确识别高光谱遥感图像上的地物,使用交叉相关光谱匹配方法获取光谱整体形态的最大偏移量,通过对比分析吸收谷位置变化获取吸收谷位置的最大偏移量,在此基础上确定吸收谷位置的偏移范围;然后使用穷举法,进行基于整体形态和偏移范围内局部吸收谷位置的识别。在云南中甸普朗斑岩铜矿区的高光谱遥感图像的实际应用表明,该方法的识别结果比仅考虑整体形态的识别结果具有更高的可靠性,能进一步证实识别结果的准确性。

To identify objects accurately in hyperspectral imagery,we use the method of cross correlogram spectral matching（CCSM） to acquire the maximal overall offset,and analyse absorption-band positions（corresponding to the same material） to acquire the maximal position offset,then get offset range of absorption-band positions on the basis of the maximal overall offset and the maximal position offset,finally identify objects from spectral overall shape and local absorption-band position within the offset range.A case study was conducted in Pulang porphyry copper deposit,Zhongdian county,Yunnan,China.The results indicate that our methods can get more accurate identification results than the approach which considers spectral overall shape only.