Research On Compression And Reconstruction Method Of Coal Mine Surveillance Video In Emergency Environment

Underground coal mines are frequent occurrences of natural disasters.When safety accidents occur,the paralysis of power,network and other infrastructures makes it difficult for traditional video surveillance systems to perform their due role,which makes it impossible to obtain disaster information in time and delays rescue work.Therefore,it is of great significance to carry out research on the compression and reconstruction of coal mine surveillance video in emergency environment.Video data collection in emergency monitoring systems is usually performed by wireless video sensor nodes.These nodes have limited storage capacity and transmission bandwidth,and cannot directly collect and transmit video data,so data compression processing is required.Compared with traditional video compression methods,distributed video compressed sensing combines the "independent coding and joint decoding" of distributed video coding and the low coding complexity of compressed sensing,and is suitable as a compression and reconstruction framework for coal mine surveillance video.Based on the theoretical basis of distributed video compression sensing,this paper conducts in-depth research on the compression and reconstruction of coal mine surveillance video in emergency environment.The main research contents and innovation results are as follows:(1)The research and application status of distributed video compression sensing are reviewed,and the main research ideas are given,focusing on the problems of lowcomplexity compression,high-quality reconstruction and application that need to be solved urgently in the coal mine monitoring system under emergency environment.(2)Taking advantage of the Bhattacharyya coefficient to distinguish image similarity,an adaptive compression algorithm based on normalized Bhattacharyya coefficient(NBCAC)is proposed,which is effective under the premise of ensuring low computational complexity at the encoding end.Discover and reduce the spatio-temporal redundancy in coal mine surveillance video,improve the compression effect,and provide feasibility for high-quality reconstruction at low sampling rates.The simulation experiment results show that the algorithm is better than BCS-SPL and SA-BCS-SPLED in reconstruction quality and reconstruction efficiency.(3)Aiming at the problem of high-quality real-time reconstruction of coal mine surveillance video sequences,a hybrid multi-hypothesis residuals reconstruction(NBCAC-MHRR)algorithm based on the contribution of side information was proposed based on the non-uniform sampling characteristics of NBCAC algorithm,which solved the problem of low overall reconstruction efficiency caused by the uneven contribution of side information provided by reference frame in multi-hypothesis prediction to a certain extent.Simulation experiment results show that the algorithm is superior to MH-BCS-SPL,MR-MHRR and HEVC-ME in reconstruction quality,and the reconstruction efficiency at low sampling rates is significantly improved.(4)Combined with the requirements of the reconstruction quality and real-time performance of the coal mine video monitoring system in emergency environment,the reference frame selection scheme in the multi-hypothesis prediction framework is optimized,and the NBCAC-MHRR algorithm is applied to the compression and reconstruction of the coal mine monitoring video,which provides a useful reference for the theoretical research to the practical application.The paper has 42 pictures,9 tables,and 82 references.