Research On Image Processing Algorithms For Power Equipment Based On Compressed Sensing

With the development of information technology and the introduction of Internet of Things technology in the process of safe production and transmission of power equipment,the amount of image monitoring data collected by various heterogeneous and complex sensors has increased geometrically.A large number of data has seriously impacted the existing equipment of monitoring,transmission and storage of power equipment.Compressed sensing theory breaks the traditional way of thinking in signal collection and analog-to-digital conversion.In this paper,the basic theory and concept of compressed sensing are introduced.The existing construction methods of measurement matrices are analyzed in detail.The advantages and disadvantages of each method are simulated and analyzed.A new measurement matrix is constructed and applied to wavelet transform and block compressed sensing.Two different image processing algorithms are proposed.This paper mainly does the following work:(1)In view of the large amount of image data transmission,node energy consumption,equipment life,transmission and storage equipment constraints in the process of power grid security monitoring,a compressed sensing image processing algorithm is proposed.At the beginning of this paper,the basic theory and concepts of compressed sensing are introduced.The construction requirements of measurement matrix and the commonly used reconstruction algorithms are analyzed in detail.The optimization of measurement matrix based on singular value decomposition is proposed.Through experimental simulation,the peak signal-to-noise ratio(PSNR) of the optimized measurement matrix is 1-2 dB higher than that of the unoptimized one.(2)Comparing with the classical compressed sensing image processing algorithm,a new compressed sensing image processing algorithm based on wavelet transform is proposed,which combines the wavelet transform and compressed sensing theory.The measurement matrix based on chaotic system worms population sequence is constructed and applied to the wavelet transform,and the experimental simulation is carried out.(3)Aiming at the complexity of the whole sampling computation of compressed sensing,the idea of block compressed sensing is selected.An image processing(4)algorithm based on block compressed sensing is proposed.A measurement matrix based on Cat sequence of chaotic system is constructed and applied to block compressed sensing,and the experimental simulation is carried out.(5)Aiming at the wireless image transmission of video and image data monitored by power equipment,the main hardware and software operation steps of wireless sensor network are introduced respectively,the perception and coordination of nodes are designed,and the overall hardware and software implementation design is carried out.Wireless image sensor network is an important part of wireless sensor network.Its characteristics and application means are basically the same as those of wireless sensor network.(6)Select the simulation software MATLAB 7.0,which is commonly used in laboratory,establish the simulation environment of outdoor power equipment monitoring,design simulation experiments,and apply the above two image processing algorithms at the same sampling rate.After experimental verification,the above image processing algorithms optimize the effect of image processing to a certain extent.Figure [57] Table [4] Reference [65].