矿井复杂环境视频监控图像增强算法研究  

Image enhancement algorithm for video surveillance in complex environment of underground mine

作　　者：王诚聪 刘亚静 WANG Cheng-cong;LIU Ya-jing(College of Mining Engineering,North China University of Technology,Tangshan 063210,China)

机构地区：[1]华北理工大学矿业工程学院,河北唐山063210

出　　处：《煤炭工程》2021年第4期147-151,共5页Coal Engineering

摘　　要：针对井下环境亮度低,粉尘严重等造成监控视频图像光照不均匀以及模糊不清等问题,提出了基于二维伽马函数实现井下视频图像增强的算法。该算法运用具有边缘保持性的引导滤波提取光照分量,通过二维伽马函数自适应调整光照,并运用基于受限自适应的直方图均衡化调整全图的对比度,从而提高图像的清晰度以及信息量。提出的算法与经典算法相比,无论是在视觉效果上,还是在信息熵,平均梯度,标准差等方面都优于经典算法。结果表明该方法可有效提高图像的清晰度,信息量以及对比度,同时减轻井下光照不均匀以及粉尘造成的图像质量不高的问题,提高了井下视频监控图像中的整体的视觉效果。Aiming at the uneven brightness and blurry problem of surveillance video images caused by the low brightness and thick dust in the underground mines,an algorithm to enhance the underground video image based on two-dimensional gamma function is proposed. The algorithm uses edge-preserving guided filtering to extract the light components,adjusts the light through the two-dimensional gamma function to achieve uneven illumination adjustment,and uses the restricted adaptive histogram equalization to adjust the contrast of the entire image,thereby improving the clarity and amount of information of the surveillance video. The proposed algorithm can combine the advantages of the retinex and gamma algorithms,and is superior to the above two algorithms in terms of information entropy,average gradient,and standard deviation. The results show that it can effectively improve the clarity,amount of information and contrast of the image,and at the same time alleviate the uneven lighting and dust blurred image in the underground mines,and improve the overall visual effect of the underground video surveillance.

关 键 词：引导滤波 二维伽马函数 视频监控 图像增强 

分 类 号：TD76[矿业工程—矿井通风与安全]