Microwave radiation image reconstruction based on combined TV and Haar Basis

摘要

Due to the complicated structure of microwave radiometric imaging system and the massive amount of data collection in one snapshot, it is difficult to achieve good performance by interferometry based on the Nyquist sampling and conventional microwave radiation imaging method. In this paper, we use the random observation matrix to sparsely sample microwave radiation image on the basis of digging compressible information of microwave radiation image, reducing the amount of data collection. Considering that sparseness of microwave radiation image on traditional sparse basis (such as: TV and Haar wavelet sparse basis) can't meet the requirements, we combine TV and Haar wavelet sparse basis to sparsely represent microwave radiation image, adopt the method of OMP algorithm finding the optimal atom to reconstruct the original microwave radiation image. We use synthesis data to simulate the three methods of combined TV and Haar wavelet, TV, Haar wavelet. The simulation results show that the proposed method to reconstruct microwave radiation image is better than the reconstruction method of single orthogonal basis.