煤矿井下视频多目标轨迹跟踪方法研究与应用

发布时间：2018-03-18 11:16

本文选题：煤矿井下　切入点：视频处理　出处：《西安科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：煤炭是我国的基础能源之一,是国家经济的重要支柱。矿难事故的频发引起了国家和社会对煤矿安全问题的高度重视。随着视频监控技术的快速发展,视频监控技术已在煤炭安全生产监测中得到广泛应用。论文从安全生产的角度,对煤矿井下的视频监控技术进行研究,通过计算机信息化技术减少煤矿井下事故的发生,保障煤矿井下的安全生产。论文根据煤矿井下监控视频的特点,研究了煤矿井下视频处理技术,对CamShift算法进行了改进,开发了煤矿井下视频多目标轨迹跟踪系统。论文的主要研究内容及创新点如下:(1)提出了一种适用于煤矿井下的视频多目标轨迹跟踪算法。为了提高煤矿井下视频目标跟踪的准确率,对运动目标进行有效跟踪,论文将小波变换和背景差分法相结合,对CamShift算法进行改进。首先采用小波三层变换对视频图像进行去噪处理,然后再进行背景差分运算,检测出运动目标,最后采用CamShift算法对运动目标进行跟踪处理。实验结果表明:改进的CamShift算法减少了原始CamShift算法在初始候选目标时的随机性,提高了目标检测和跟踪的准确率。(2)开发了煤矿井下视频多目标轨迹跟踪系统。为了应对煤矿井下环境多变的情况,研究了视频处理技术和煤矿井下视频的特点,将多种目标检测技术和目标跟踪技术融入到一起应用于煤矿井下的视频中,开发了煤矿井下视频多目标轨迹跟踪系统。在系统开发过程中,采用模块化的设计思想,将各个功能独立封装,方便用户选择和使用。并且系统还提供了危险区域警告功能,当运动目标进入到人工设置的危险区域时,该目标将会被突出显示,以示警告。论文开发的煤矿井下视频多目标轨迹跟踪系统运用了改进的CamShift算法,提高了目标跟踪的成功率。但系统中仍然存在一些问题,今后将进一步进行研究和探索。
[Abstract]:Coal is one of the basic energy sources in our country and an important pillar of the national economy. The frequent occurrence of mine accidents has caused the country and society to attach great importance to the problem of coal mine safety. With the rapid development of video surveillance technology, Video surveillance technology has been widely used in coal mine safety monitoring. From the point of view of safety production, this paper studies the video monitoring technology in coal mine, and reduces the occurrence of mine accidents by computer information technology. According to the characteristics of monitoring video in coal mine, the video processing technology of underground coal mine is studied, and the CamShift algorithm is improved. A video multi-target trajectory tracking system is developed. The main contents and innovations of this paper are as follows: 1) A video multi-target trajectory tracking algorithm suitable for coal mine is proposed. In order to improve the mine video trajectory tracking algorithm, the main contents and innovations of this paper are as follows: (1) A video multi-target trajectory tracking algorithm is proposed. Target tracking accuracy, In this paper, CamShift algorithm is improved by combining wavelet transform with background difference method. Firstly, wavelet three-layer transform is used to Denoise the video image, then background difference operation is carried out. The moving target is detected and the moving target is tracked by CamShift algorithm. The experimental results show that the improved CamShift algorithm reduces the randomness of the original CamShift algorithm in the initial candidate target. In order to deal with the changing environment of underground coal mine, the video processing technology and the characteristics of underground video are studied in order to improve the accuracy of target detection and tracking. The multiple target detection technology and target tracking technology are integrated into the video of underground coal mine, and a multi-target tracking system is developed. In the course of the system development, the modular design idea is adopted. The system also provides the danger area warning function, which will be highlighted when the moving target enters the manually set danger area. A mine video multi-target trajectory tracking system developed in this paper uses the improved CamShift algorithm to improve the success rate of target tracking. However, there are still some problems in the system, which will be further studied and explored in the future.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN948.6;TP391.41

【参考文献】