关键词: 矿井机电设备图像, Shearlet变换, 均值, 标准差, 图像块, 维纳滤波, 阈值去噪函数模型

Abstract: The operation of mine electrical and mechanical equipment with high efficiency is an important quarantee to ensure mine safety production.Due to the existing factors such underground dust with high density and uneven illumination,the electrical and mechanical equipment images obtained by the video image monitoring system are obscure relatively,therefore,the monitoring effects of the mine electrical and mechanical equipment are influenced.A new mine electrical and mechanical equipment video monitoring images are proposed based on combing with Shearlet transform and adaptive classification method of image region.Firstly,combing with the mean and standard deviation of the local regions of image,a new adaptive classification method image regions is designed,the image is divided into homogeneous image blocks,non-homogeneous image blocks and edge image blocks,the homogeneous image blocks are processed by Wiener filtering algorithm;secondly,the non-homogeneous image blocks are conduct multi-scale Shearlet transform,the low-frequency and high-frequency Shearlet decomposition coefficients are obtained,a new Shearlet transform domain adaptive threshold denoising function model is put farward to filtering the noise in high-frequency Shearlet decomposition coefficients,the original low-frequency Shearlet decomposition coefficient and filtered high-frequency Shearlet decomposition coefficients are conducted decomposition coefficients reconstruction;finally,the edge image blocks,filtered homogeneous image blocks and non-homogeneous image blocks are superimposed.Programmes of the Wiener filtering algorithm,hard wavelet threshold denoising function model,soft wavelet threshold denoising function model and the algorithm proposed in this paper are written based on MATLAB language,the experimental result show that the performance of the algorithm proposed in this paper is better than other algorithms,and the visual effects of the images proposed by it is improved greatly,it has some reference for improving the quality of the mine electrical and mechanical equipment video monitoring images.

Key words: Mine electrical and mechanical images, Shearlet transform, Mean, Standard deviation, Image blocks, Wiener filtering, Threshold denoising function model