Of Impulse Gpr Based On Compressed Sensing Imaging Studies

Compressed Sensing (CS) theory is a novel data collection and coding theory under the condition that signal is sparse or compressible. It first employs nonadaptive linear projection to preserve the structure of the original signal information, and then the accurately reconstruction of original signal from by solving numerical optimization problems. This makes the compressed sample data less than the tradition theory of the Shannon-Nyquist sampling. It has broad applications such as compressive imaging, image and video processing systems, Radar imaging, MRI imaging, etc.The paper mainly studies the application of the CS theory in the pulse GPR imaging. Firstly, we begin to study the measurement matrix of the CS theory based on the Chinese remainder theorem, to construct a certainty the conditions of CS theory RIP measurement matrix-Sun measurement matrix. The matrix generalies processes on the basis of the polynomial measurement matrix generalies principle, the performance of random measurement matrix and deterministic measurement matrix were compared, obtained in the same measured number, the sun matrix measurement error is smaller than the randomness of the measurement matrix reconstruction and the hardware implementation more simple.Secondly, the pulse GPR signal reconstruction algorithms were researched. Through the study of the classic CS reconstruction algorithm, the random pulse GPR echo signal imaging aperture-minimum {1norm imaging algorithm is mentioned and compare signal reconstruction algorithm, The imaging algorithm is put forward in this paper the measurement data less, Underground targets required for image reconstruction data than the Shannon-Nyquist sampling at least one order of magnitude less computation and underground targets of the reconstructed image is more accurate.Finally, we construct the Sun measurement matrix and the random aperture-minimum {1-norm algorithm for underground targets MATLAB imaging simulation and csuGPR of data processing software the measured radar data. Imaging of the measured data compared to the effect of recursive back projection imaging algorithm with the classical least squares method of imaging algorithms, random aperture measured by CS measurement random aperture-minimum {1-norm imaging to obtain the sparse target space image only a few clutter echo, little effect on the imaging effect but makes data acquisition is greatly reduced and is robust to the noise.