基于IDL的HSI图像条带噪声去除方案研究

Study On Destriping Methods of HSI Images Based on IDL

HJ-1-A卫星超光谱数据的绿、红和红外各波段(21-115波段)数据质量较高,具有良好的应用潜力.但蓝光波段(前20个波段)有明显的条带噪声,对超光谱数据信息提取的精度产生了严重影响,需去除噪声才能使用.针对HJ-1-A星HSI图像条带噪声的特殊性,提出了两种条带去除方案,并给出用IDL语言实现的方法.结果表明,该方法能够在保持原始HSI图像特征的前提下有效地去除条带噪声.

In HJ-1-A satellite HSI data, the spectrums of the green, red and infrared （21-115 bands） are clear and have a good application potential. However, in the blue spectrum （bands before 20）, the strips are obviously and have a serious impact on the information extraction. In order to use the blue spectrum, the strips should be eliminated. Considering the specialities of HSI data, two methods were proposed and programmed by IDL. Experimental result shows that the methods can effectively eliminate the strips under the premise of maintaining the original HSI image features.