论文平台收录红外图像增强AlgorithmNOAA-20radiance retrievalreflectance retrievalshort-wave infrared (SWIR) bandsstriping mitigationvisible infrared imaging radiometer suite (VIIRS)
Earth View Reflectance Image Striping Mitigation for NOAA-20 VIIRS Short-Wave Infrared Bands
NOAA-20 VIIRS短波红外波段地球观测反射率图像的条带抑制
Ning Lei; Xiaoxiong Xiong; Kevin Twedt; Sherry Li; Xu Geng
- Journal / Publication
- IEEE Transactions on Geoscience and Remote Sensing
- Published
- 时间暂未填写
论文摘要
The visible infrared imaging radiometer suite (VIIRS) is an Earth-observing satellite sensor. Out of the 22 VIIRS spectral bands, five are the short-wave infrared (SWIR) bands. Earth views of these bands for the VIIRS instrument aboard the NOAA-20 (N20) satellite have unexpected striping. The scene spectral radiance is retrieved by using a cubic polynomial of the background removed detector digital number with the offset set to zero. To minimize the striping, we develop an algorithm to find the offsets from the striping assessments of the satellite images of the Sahara Desert, deep convective clouds (DCCs) around the equator, and the Moon. We use the average of the offsets obtained from these three scenes and the other coefficients in the original cubic polynomial to update the polynomial coefficient lookup table that is used for the NASA N20 VIIRS top-of-the-atmosphere (TOA) spectral radiance and reflectance retrievals in Collection 2.1. After the application of the derived offsets, the striping from the scenes and the Libya 4 Desert is significantly reduced, except bands I3 and M10 over the DCC scene. For the DCC scene, the original striping for bands I3 and M10 is reduced but not significantly. In the case that the quadratic term in the polynomial also contributes to the observed striping, we further develop an algorithm to find both the offset and the quadratic term coefficient from striping assessments of two scenes whose spectral radiance levels differ significantly.
