Research On Image Reconstruction Algorithm Based On Compressed Sensing In Underground Coal Mine

In recent years,the Internet of Things technology has been introduced into coal mine safety production.The real-time image data of underground coal mines obtained through image sensing technology is increasing day by day.The existing information transmission and data storage equipment in underground coal mines are facing unprecedented challenges.Nyquist sampling theorem is limited by 2 times the bandwidth width,and compressive sensing theory breaks the bandwidth limitation with sparse characteristics.In this paper,two aspects of signal reconstruction algorithm and signal sparse representation in compressed sensing theory are studied in depth.The basic work includes the following two aspects:(1)In-depth research on the matching and tracking algorithms in the greedy iterative algorithm,analysis of its theory,and based on this,an improved algorithm based on the OMP algorithm-sparsity adaptive piecewise orthogonal matching tracking Adaptive Stagewise Orthogonal Matching Pursuit(SAStOMP)algorithm.This algorithm combines adaptive thinking,variable step size iterative thinking,and piecewise orthogonal thinking.Under the condition of unknown signal sparsity,adaptively selecting the number of atoms in the support set can achieve accurate signal to a certain extent.(2)Based on the study of sparse representation of signals,a wavelet transform algorithm and compressed sensing theory are combined to propose an improved method of compressed sensing image processing based on wavelet transform.Sym8 wavelet was selected as the sparse basis to multi-layer decompose the coal mine underground image,and then it was reconstructed and restored using the improved SAStOMP algorithm.The transmitted lowfrequency coefficients are reconstructed by inverse wavelet transform.The MATLAB platform was selected for simulation experiments,and the effect of reconstruction was evaluated by comparing the signal-to-noise ratio of image signal reconstruction under different algorithms at the same sampling rate.The algorithm simulation proves that the algorithm in this paper can reconstruct the original image better than the classic compressed sensing reconstruction algorithm,and has practical value in the application of the actual coal mine image.