欢迎访问《金属矿山》杂志官方网站,今天是 分享到:

金属矿山 ›› 2017, Vol. 46 ›› Issue (07): 151-154.

• 机电与自动化 • 上一篇    下一篇

小波变换域井下视频监控图像改进阈值去噪方法

申红   

  1. 吉林省经济管理干部学院计算机实验中心,吉林 长春 130012
  • 出版日期:2017-07-15 发布日期:2017-09-13

Improved Threshold Denoising Method of Mine Video Monitoring Image in Wavelet Transform Domain

Sheng Hong   

  1. Computer Experiment Center,Jilin Province Economic Management Cadre College,Changchun 130012,China
  • Online:2017-07-15 Published:2017-09-13

摘要: 通过对井下视频监控系统实时采集的各类开采、地质环境信息进行有效分析,可为制定井下开采方案以及进行灾害救援提供准确依据。由于井下光照不均匀,粉尘较多,严重干扰了监控探头获取高质量的监控图像。为此,结合小波变换方法,提出了一种小波变换域改进阈值去噪方法。该方法首先对井下视频监控图像进行3层小波分解,得到低频和高频分解系数;其次对低频分解系数进行重构,得到空间域背景图像,采用维纳滤波算法进行处理,以去除其中存在的少量噪声;然后根据经典小波硬、软阈值去噪模型的不足,提出了一种改进型阈值去噪模型,该模型可分别根据不同的高频分解系数自适应设定阈值,可有效去除不同分解层高频系数中的噪声,对去噪后的各高频分解系数进行重构,得到空间域细节图像;最后,分别将去噪后的空间域背景图像和细节图像进行叠加,得到去噪后的井下视频监控图像。采用1幅内蒙古某煤矿井下视频监控图像进行试验,并引入了经典小波硬、软阈值去噪模型及2类已有的改进型阈值去噪模型与所提方法进行试验对比,结果表明,所提方法不仅可有效去除井下视频监控图像中的噪声,而且可保持图像细节信息完整性。

关键词: 井下视频监控系统, 小波变换, 背景图像, 维纳滤波, 阈值去噪模型, 细节图像

Abstract: Analyzing the all kings of mining and geological environmental information obtained by mine video monitering sysyem effectively can provide the accurate basis for the formulation of underground mining scheme and disaster assistance.Due to the existing factors of uneven illumination and high concentration of dust,the acquistion of the high quality images of the mine video monitoring sysyems are affected seriously.Combing with wavelet transform,a new improved threshold denoising method in wavelet transform domain is proposed.Firstly,the mine video monitoring image is conducted three-layers wavelet decomposition,the low-frequency wavelet decomposition coefficients and high-frequency wavelet decomposition coefficients are obtained;then,the background image in spatial domain is acquired by conducting low-frequency wavelet decomposition coefficients reconstruction,besides that,the Wiener filtering algorithm is used to deal with the small amount noises existed in the background image in spatial domain;thirdly,according to the deficiencies of the classical wavelet hard and soft threshold denoising models,a new improved threshold denoising model is put forword,the thresholds of the improved denoising model can be obtained based on the different of the high-frequency wavelet decomposition coefficients,it is used to filter out the noises in the high-frequency wavelet decomposition coefficients of different layers of wavelet decomposition,the filtered high-frequency wavelet decomposition coefficients are conducted coefficients reconstruction,so the detailed image in spatial domain is get;finally,the filtered background image in spatial domain and the filtered detailed image in spatial domain are superimposed to obtained the mine video monitoring image with high definition.The experimental data of the one mine video monitoring images acquired from a coal mine in Inner Mongolia,the experimental results show that the denoising effects of the denoising method proposed in this paper is superior to the classical wavelet hard and soft denoising models and two existed improved wavelet threshold denoising models.The experimental results further show that the denoising method proposed in this papaer can provide some reference for improving the quality of mine video monitoring images.

Key words: Mine video monitoring system, Wavelet transform, Background image, Wiener filtering, Threshold denoising model, Detailed image