A Microwave Radiation Interferometry Method Based on Adaptive Super-sparse Sampling
The Interferometry Synthetic Aperture Imaging Radiometers (SAIRs) is to sample visibility function based on the Nyquist theory of space interferometry measurement, which do not need the mechanical scanning and can directly image. Due to the complex structure of the imaging system and low imaging resolution, the SAIRs practical application is limited seriously. According to the characteristic of the image is sparse or can be sparse representation in transform domain, the Compressed Sensing (CS) can project the high-dimensional signal to low-dimensional space, so the quantity of the projection measurement data is far less than that by the Nyquist sampling method. Also the microwave radiation interferometry conducted in the frequency domain, which has the characteristics of low frequency information less and high frequency information richer, and the distribution of them is centralized; at the same time, the microwave radiation image itself have the specialty of the gradient sparsity and the local smooth-ness, it can be sparse representation in differential domain. On the basis of the priori information about the observation and the sparse domain, we establish the incoherent optimization model between the observation matrix and the sparse matrix according to the principle of the two matrixes satisfying the irrelevant in the CS. Using the incoherent optimization model, we can adaptively obtain the spatial measurement with different probability, to realize super sparse interferometry. The adaptive super-sparse sampling method can overcome the disadvantage of equal probability of the Fourier random sampling methods. In order to reconstruct the microwave radiation image, we establish the imaging model based on total variation regularization constraint, and use the alternating iterative algorithm to realize the reconstruction. The simulation and experiment results show that it is fast to reconstruct microwave radiation image with the adaptive super-sparse sampling method, and it can greatly improve the quality of the microwave radiation image in the case of the same sampling rate.
Suhua Chen Lu Zhu Yuanyuan Liu
School of Information Engineering, East China Jiaotong University, Nanchang 330013, China
国际会议
Progress in Electromagnetics Research Symposium 2014(2014年电磁学研究新进展学术研讨会)
广州
英文
449-453
2014-08-01（万方平台首次上网日期，不代表论文的发表时间）