视觉计算在煤矿巷道变形监测中的应用研究
发布时间：2019-04-24 20:32
【摘要】：煤炭工业是我国最重要的基础能源产业,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。本文主要研究了立体视觉测量中图像特征点提取、摄像机标定和重建点坐标求解三个关键部分,并实验应用立体视觉测量方法监测煤矿巷道形变。主要研究内容和成果如下:首先研究了图像上角点和光斑中心像点的提取方法。实验对比了Harris、Shi-Tomasi和FAST三种角点检测算法,结果表明:Harris算法对阈值敏感且易产生角点聚簇;FAST算法易提取到伪角点和产生角点聚簇;Shi-Tomasi算法能使检测到角点分布均匀,避免角点聚簇,得出Shi-Tomasi算法适合十字靶标角点像素级坐标提取。通过图像预处理由Canny边缘检测算法得到光斑单像素边缘轮廓,利用最小二乘拟合法获取光斑中心像点坐标。其次对摄像机平面模板标定法进行了研究。由于通常计算单应性矩阵时使用了畸变较大的像点影响了内部和外部参数初值,从而影响了标定精度。本文提出在计算单应性矩阵时仅使用图像中心附近像点,使内部参数和外部参数初值能更好地逼近准确值,提高摄像机标定精度。实验证明此方法能提高摄像机标定精度。然后针对立体视觉空间点坐标求解问题,分析比对了三种常用求解方法:最小二乘法、公垂线段中点法、基于极线约束的重建方法,给出了以公垂线段为约束使反投影像点误差最小的空间点坐标求解方法,实验证明该方法能提高空间点坐标求解精度。最后将立体视觉测量技术应用到巷道变形监测中,给出了一种巷道变形监测实施方案,实验对巷道上监测点位移进行了测量,结果证明该视觉测量方案能准确计算出巷道上监测点空间位移,能对巷道变形进行监测。
[Abstract]:Coal industry is the most important basic energy industry in our country. The frequent occurrence of mine accidents has caused the state and society to attach great importance to the safety of coal mines. With the rapid development of video surveillance technology, video surveillance technology has been widely used in coal safety production monitoring. In this paper, the extraction of image feature points, camera calibration and the solution of reconstruction point coordinates in stereo vision measurement are studied, and the method of stereo vision measurement is used to monitor the deformation of coal mine roadway. The main contents and achievements are as follows: firstly, the extraction method of corner and spot center image is studied. Three corner detection algorithms, Harris,Shi-Tomasi and FAST, are compared. The results show that Harris algorithm is sensitive to threshold and it is easy to generate corner cluster, FAST algorithm is easy to extract pseudo-corner and generate corner cluster, and FAST algorithm is easy to extract pseudo-corner and generate corner cluster. The Shi-Tomasi algorithm can make the corner distribution uniform and avoid the clustering of corner points. It is concluded that the Shi-Tomasi algorithm is suitable for pixel-level coordinate extraction of cross-target corner points. The single pixel edge contour of the spot is obtained by the Canny edge detection algorithm, and the coordinates of the spot center image are obtained by the least square fitting method. Secondly, the camera plane template calibration method is studied. Since the large distortion image points are usually used in the calculation of the monotonic matrix, the initial values of internal and external parameters are affected and the calibration accuracy is affected. In this paper, it is proposed to use only image points near the image center in calculating the homomorphism matrix, so that the initial values of internal and external parameters can be better approximated to the exact values and the camera calibration accuracy can be improved. Experiments show that this method can improve camera calibration accuracy. Then, aiming at the point coordinate problem of stereo vision space, three common methods are analyzed and compared: least square method, middle point method of common vertical segment, and reconstruction method based on polar constraint. This paper presents a method for solving spatial point coordinates with the constraint of common vertical segment to minimize the error of back projection image points. The experimental results show that this method can improve the accuracy of spatial point coordinates. Finally, the stereo vision measurement technology is applied to the tunnel deformation monitoring, and a kind of tunnel deformation monitoring implementation scheme is given. The displacement of the monitoring points on the roadway is measured by experiments. The results show that the visual measurement scheme can accurately calculate the spatial displacement of the monitoring points on the roadway and can monitor the deformation of the roadway.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD326;TP391.41
本文编号：2464762
[Abstract]:Coal industry is the most important basic energy industry in our country. The frequent occurrence of mine accidents has caused the state and society to attach great importance to the safety of coal mines. With the rapid development of video surveillance technology, video surveillance technology has been widely used in coal safety production monitoring. In this paper, the extraction of image feature points, camera calibration and the solution of reconstruction point coordinates in stereo vision measurement are studied, and the method of stereo vision measurement is used to monitor the deformation of coal mine roadway. The main contents and achievements are as follows: firstly, the extraction method of corner and spot center image is studied. Three corner detection algorithms, Harris,Shi-Tomasi and FAST, are compared. The results show that Harris algorithm is sensitive to threshold and it is easy to generate corner cluster, FAST algorithm is easy to extract pseudo-corner and generate corner cluster, and FAST algorithm is easy to extract pseudo-corner and generate corner cluster. The Shi-Tomasi algorithm can make the corner distribution uniform and avoid the clustering of corner points. It is concluded that the Shi-Tomasi algorithm is suitable for pixel-level coordinate extraction of cross-target corner points. The single pixel edge contour of the spot is obtained by the Canny edge detection algorithm, and the coordinates of the spot center image are obtained by the least square fitting method. Secondly, the camera plane template calibration method is studied. Since the large distortion image points are usually used in the calculation of the monotonic matrix, the initial values of internal and external parameters are affected and the calibration accuracy is affected. In this paper, it is proposed to use only image points near the image center in calculating the homomorphism matrix, so that the initial values of internal and external parameters can be better approximated to the exact values and the camera calibration accuracy can be improved. Experiments show that this method can improve camera calibration accuracy. Then, aiming at the point coordinate problem of stereo vision space, three common methods are analyzed and compared: least square method, middle point method of common vertical segment, and reconstruction method based on polar constraint. This paper presents a method for solving spatial point coordinates with the constraint of common vertical segment to minimize the error of back projection image points. The experimental results show that this method can improve the accuracy of spatial point coordinates. Finally, the stereo vision measurement technology is applied to the tunnel deformation monitoring, and a kind of tunnel deformation monitoring implementation scheme is given. The displacement of the monitoring points on the roadway is measured by experiments. The results show that the visual measurement scheme can accurately calculate the spatial displacement of the monitoring points on the roadway and can monitor the deformation of the roadway.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD326;TP391.41
【引证文献】
相关期刊论文 前1条
1 张俊义;;近距离测量在煤矿巷道变形监测中的应用研究[J];煤炭与化工;2017年09期
，本文编号：2464762
本文链接：https://www.wllwen.com/kejilunwen/kuangye/2464762.html
