Research On Wavelet Image Denoising And Compressing Method For Mobile Internet Of Things

With the advent of the Mobile Internet of Things, people are becoming more and moreparticular about the transmission of image information in the network. They not only ask forhigh image quality, but also high transmission speed. The main factor affecting the imagequality is noise. And images contain a substantial amount of data. In order to achieve highspeed, the images need to be compressed before transmission and without affecting the imagequality.After analysis the characteristics of the Mobile Internet of Things, this paper proposes anew wavelet shrinkage image denoising algorithm based on spherical coordinate and animproved image compression method.The new algorithm of image denoising is improved as follows:(1) A new shrinkage threshold in spherical coordinate domain is proposed. This newthreshold does not require pre-estimated the image noise variance and reduced the complexityof the algorithm.(2) A new non-linear adaptive shrinkage function is proposed in this paper. The newfunction is continuous at the threshold and approaches true value of the wavelet coefficients ata high speed. The new function can enhance the edge and contour features of the images. It iscontrolled by Multi-Scale Model Product (MSMP) defined in this paper, which caneffectively sepatate the image information and noise information.An improved image compression algorithm is introduced from the following parts:(1) Low-frequency part is processed by DPCM predictive coding. Use the coefficients onthe middle row and middle column to predict the unkown wavelet coefficients.(2) High-frequency part is processed by improved SPIHT coding. First, the waveletcoefficients are pre-treated, then, a Matrix of Maximum Pixel (MMP) is introduced to avoidthe duplication scans of the coefficients in the traditional SPIHT algorithm. The improvedalgorithm can achieve higher compression ratio and need fewer comparison.Finally, Matlab simulation software is employed to conduct simulation. Experimentalresults compared with similar algorithms show that the improved image denoising methodand image compression algorithm have good results. The new methods proposed in this paperare practical image processing methods.