分类: 地球科学 >> 空间物理学 提交时间: 2017-03-10
摘要: Dual-Frequency Polarized Scatterometer (DFPSCAT) is a pencil-beam rotating scatterometer which is used to measure snow water equivalence (SWE). Respecting the low azimuth resolution of its forward-looking region, an adaptive regularization deconvolution super-resolution method, based on the scatterometer echo signal model, is proposed. Compared with the classical SIR and MAP algorithms, the proposed method can better reconstruct the original signal, and has less noise amplification. The algorithm processing accuracy with different Kpcis also studied, and the results show that when the value of Kpcis less than 0.1, nearly the entire restored data can satisfy the requirement of 0.5dB accuracy.
分类: 地球科学 >> 空间物理学 提交时间: 2017-03-10
摘要: Dual-Frequency Polarized Scatterometer (DFPSCAT) is a new system utilizing Doppler beam sharpening (DBS) technology for azimuthal resolution enhancement. Considering the DBS technology is inapplicable for the middle areas of the swath, a theoretical framework of deconvolution signal processing is proposed to improve resolution. A deconvolution method of the nonlocally centralized sparse representation (NCSR) is adopted to verify its feasibility, and simulation results show that the deconvolution method have obviously better resolution enhancement and higher recovery accuracy than these of the classical scatterometer image reconstruction (SIR) method. �2016 IEEE.
分类: 地球科学 >> 空间物理学 提交时间: 2016-05-03
摘要: Dual-Frequency Polarized Scatterometer (DFPSCAT) is a pencil-beam rotating scatterometer which is used to measure snow water equivalence (SWE). Respecting the low azimuth resolution of its forward-looking region, an adaptive regularization deconvolution super-resolution method, based on the scatterometer echo signal model, is proposed. Compared with the classical SIR and MAP algorithms, the proposed method can better reconstruct the original signal, and has less noise amplification. The algorithm processing accuracy with different K-pc is also studied, and the results show that when the value of K-pc is less than 0.1, nearly the entire restored data can satisfy the requirement of 0.5dB accuracy.