矿山远程监控图像的改进非局部均值滤波算法  

Improved Non-Local Means Filtering Algorithm of Mine Remote Monitoring Image

作　　者：李龙海 郭华锋[1] 于萍[1] 刘磊[1] 陆兴华[1] LI Longhai;GUO Huafeng;YU Ping;LIU Lei;LU Xinghua(School of Mechanical and Electrical Engineering,Xuzhou University of Technology,Xuzhou 221018,China)

机构地区：[1]徐州工程学院机电工程学院,江苏徐州221018

出　　处：《金属矿山》2022年第10期165-169,共5页Metal Mine

基　　金：江苏高校自然科学研究项目(编号:17KJB460015);江苏省自然科学基金(青年基金)项目(编号:BK20180177)。

摘　　要：“少人化”“无人化”综采工作面是煤矿发展的重要方向,同时也是煤矿安全生产的重要保障。矿山远程监控系统则是其中的重要一环,由于采煤工作面环境复杂,光线昏暗、高粉尘、成像电压不稳定、对比度低等诸多因素严重影响了视频监控图像的质量。为提高图像质量兼顾传输效率,在经典非局部均值滤波算法的基础上提出了一种基于积分图像的快速非局部均值滤波算法。分别将所提算法与经典非局部均值滤波算法、中值滤波算法以及不同窗口尺寸的均值滤波算法进行性能对比分析,并结合主观评价法和客观评价法对各算法去噪后的图像质量进行了评价。结果表明:经典非局部均值滤波算法以及所提算法均能起到较好的降噪效果,但在运行时间上,前者耗时约为后者的40倍,难以实现图像的实时处理。所提算法不仅为图像降噪提供了新思路,而且有助于实现高效处理远程监控图像,进而实现无人化、少人化开采。"Less people"or"No people"fully-mechanized coal mining face is not only an important direction of coal mine development,but also an important guarantee for coal mine safety production.The remote monitoring system of coal mine is an important part of them.However,many factors,such as dim light,high dust,unstable imaging voltage and low contrast,seriously affect the quality of video monitoring image in the complex environment.In order to improve the image quality and transmission efficiency,a fast non-local mean filtering algorithm based on integral image is proposed based on the classical nonlocal mean filtering algorithm.On this basis,the performance of the proposed algorithm is compared with the classical non local mean filtering algorithm,median filtering algorithm and mean filtering algorithm with different window sizes.The image quality after denoising using each algorithm is evaluated by combining the subjective and objective evaluation methods.The results show that both the classical non local mean filtering algorithm and the proposed algorithm can achieve good noise reduction effect,but the running time of the former is about 40 times that of the latter,which makes it difficult to achieve real-time image processing.The proposed algorithm not only provides a new idea for image noise reduction,but also helps to achieve efficient processing of remote monitoring images,thus realizing unmanned mining and less people mining.

关 键 词：远程监控 综采自动化 非局部均值滤波 积分图 降噪 

分 类 号：TD672[矿业工程—矿山机电] TP391[自动化与计算机技术—计算机应用技术]