工作面监控视频清晰化应用技术研究
Research on clarity technology for working face monitoring videos
张杰 1郭建军 1王永刚 1高巾栋 1曹敏 1方云峰 1于继斌 1高亮亮2
作者信息
- 1. 国能包头能源有限责任公司 李家壕煤矿,内蒙古 鄂尔多斯 017000
- 2. 北京天玛智控科技股份有限公司,北京 101320
- 折叠
摘要
针对井下工作面监控视频的视觉清晰化问题,首先提出一种监控摄像仪防尘自动疏水方法,利用疏水涂层制备摄像仪视窗,有效减少煤尘、水雾等附着,提高监控视频画面的可视化监控效果.针对工作面工况条件下粉尘雾气导致监控视频图像质量下降问题,提出一种监控视频图像实时去雾算法.通过实时去雾处理后监控图像峰值信噪比保持在 28.2 左右,使得监控图像特征信息很好保持的同时,增强工作面监控视觉效果.在监控视频图像视觉畸变矫正技术研究过程中,提出一种工作面监控视频图像校正方法,充分利用视频图像内部特征信息,无需外部标定实现监控视频图像边缘畸变校正.实验表明,文章提出的工作面监控视频清晰化技术一定程度上能够满足实际生产需求,提升工作面监控视频可视化效果.
Abstract
Aiming at the problem of visual clarity of monitoring video in underground working face,we proposed a dust-proof automatic drainage method for monitoring camera.The hydrophobic coating was used to prepare the camera window,which can effectively reduce the adhesion of coal dust and water mist,and improve the visual monitoring effect of monitoring video screen.In view of the monitoring video image quality degradation caused by dust fog under working condition of working face,a real-time defogging algorithm for monitoring video image was proposed.After real-time defogging,the peak signal-to-noise ratio of the monitoring image was maintained at about 28.2,which makes the feature information of the monitoring image well maintained and enhances the visual effect of the working face monitoring.In the research process of visual distortion correction technology of surveillance video image,a correction method of surveillance video image in working face was proposed,which made full use of the internal feature information of video image and realized the edge distortion correction of surveillance video image without external calibration.Experiments show that the proposed technology can meet the actual production needs to a certain extent and improve the visualization effect of working face monitoring video.
关键词
监控视频清晰化/摄像仪防尘疏水/图像实时去雾/图像畸变校正/可视化监控Key words
monitoring video clarity/dust-proof and hydrophobic camera/image defogging/image distortion correction/visual monitoring引用本文复制引用
基金项目
天地科技股份有限公司科研项目(2021-TD-MS013)
出版年
2024
NETL
NSTL
万方数据