Research Of Image Processing Algorithm For Underground Coal Mine Based On Compression Sensing

With the development of technology information and Internet of things technology is introduced in coal mine safety production,real-time image monitoring data collected by heterogeneous complex sensors for production processes,safety environments,personnel,equipment,etc.already showing a multiple growth,massive data transmission seriously impacted the existing coal mine transmission and storage equipment Compressive Sensing breaking the traditional signal sampling transmission by 2 times the bandwidth limit,sparsely processed the image signal to other spatial domains,using non-adaptive linear projection to maintain the original image signal structure,and then capture,codec,and finally through the algorithm to reconstruct the recovery signal,by collecting a small amount of information,without distortion to reconstruct the original signal.To achieve a great savings in space storage and resource sampling.(1)Aiming at the problem of large data transmission capacity,node energy consumption and life limitation in the existing coal mine,this paper uses the compression perceptual principle and compares it with the classical compression sensing algorithm.Combining the wavelet transform algorithm with the compression perceptual theory,that an improved image processing method based on wavelet transform is proposed.(2)Select the simulation software system Matlab7.0,the establishment of simulation environment,design simulation experiments,By reconstructing the signal-to-noise ratio and reconstruction time of the reconstructed image signal of different algorithms at the same sampling rate,which used to expressing the effect of reconstruction.(3)Exploring the Realization of Improved Compression Sensing Image Processing Algorithm Based on Wavelet Transform in FPGA and Its Realization in FPGA Hardware.(4)Through the simulation of the algorithm,this algorithm can reconstruct the original image,save the resources and the energy better under the same sampling rate compared with the original wavelet transform compression sensing algorithm.