Pre-Processing of TASI Data

Mainly contains 3 parts： (1) Removal of Noise (2) Atmospheric Correction (3) Separation of Temperature and Emissivity

(1) Removal of Noise

After obtaining the noise level of each band, the image de-noising method based on sparse representation is used to remove the noise data.【1】

(2) Atmospheric Correction

A combination of atmospheric correction algorithm is employed to counter the problem of Autonomous Atmospheric Compensation (AAC) algorithm diversity of inversion results【2】, which combined with the In-Scene Atmospheric Compensation (ISAC) blackbody pixels calibration method【3】.It improved the inversion precision of the atmospheric wave spectrum and made the surface radiation brightness value more accurate.

(3) Separation of Temperature and Emissivity

the Correlation Based Temperature Emissivity Separation Algorithm (CBTES) is used for classification of temperature and emissivity【4】. The method compared the surface emissivity spectral radiation and the atmospheric downward curve correlation as the basis of the surface temperature optimization. The smaller the correlation, the higher the surface temperature valuation accuracy is.

【1】Aharon, M. , M. Elad , and A. Bruckstein . "-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation." 54.11(2006):4311-4322.

【2】 Gu, D. G. , et al. "Autonomous atmospheric compensation (AAC) of high resolution hyperspectral thermal infrared remote-sensing imagery." IEEE Transactions on Geoscience and Remote Sensing 38.6(2000):2557-2570.

【3】Stephen, et al. "An in-scene method for atmospheric compensation of thermal hyperspectral data." Journal of Geophysical Research Atmospheres (2002).

【4】 Jie, Cheng , et al. "Correlation-based temperature and emissivity separation algorithm." Science in China 3(2008):13.