基于图像处理的轨道位移监测系统研究
|
摘要
在高铁桥梁轨道伸缩调节器处的轨道位移是影响行车安全的关键薄弱点。对该处轨道的位移监测非常必要,但是在高铁线路上是不允许安装任何设备的。为了解决实时监测的需求,研究基于图像处理的位移监测系统。在轨道侧边安装固定摄像机,定时拍摄照片;使用数字图像处理算法计算出轨道的实际位移。为了达到比较好的效果,使用相位相关法求相对位移,并且使用傅里叶梅林变换解决相机振动后偏移和更换后需要重新标定变焦倍数的问题。该系统实现实时自动测量,并且取得较好的测量精度,解决非接触测量的需求。
The track displacement at the high-speed railway bridge telescopic adjuster is the key weakness affecting traffic safety. It is necessary to monitor the track displacement,but no equipment is allowed to be installed on the highspeed railway line. In order to solve the needs of real-time monitoring,we studied a displacement monitoring system based on image processing. The fixed camera was installed on the side of the track and took photos regularly. The actual displacement of the track was calculated by using the digital image processing algorithm. In order to achieve better results,the phase correlation method was used to calculate the relative displacement,and the Fourier-Mellin transform was used to solve the problem of camera deflection after vibration and the need to re-calibrate the zoom multiple after replacement. This method realizes real-time automatic measurement,and achieves better measurement accuracy,and solves the demand of noncontact measurement.
The track displacement at the high-speed railway bridge telescopic adjuster is the key weakness affecting traffic safety. It is necessary to monitor the track displacement,but no equipment is allowed to be installed on the highspeed railway line. In order to solve the needs of real-time monitoring,we studied a displacement monitoring system based on image processing. The fixed camera was installed on the side of the track and took photos regularly. The actual displacement of the track was calculated by using the digital image processing algorithm. In order to achieve better results,the phase correlation method was used to calculate the relative displacement,and the Fourier-Mellin transform was used to solve the problem of camera deflection after vibration and the need to re-calibrate the zoom multiple after replacement. This method realizes real-time automatic measurement,and achieves better measurement accuracy,and solves the demand of noncontact measurement.
引文
[1]laganiere R.Open CV2计算机视觉编程手册[M].北京:科学出版社,2013.
[2]Bradski G.学习Open CV[M].北京:清华大学出版社,2009.
[3]王胜辉,刘鹏,冯宏恩,等.基于Open CV的高压电气设备放电紫外成像检测图像量化参数提取[J].计算机应用与软件,2016,33(9):167-170,189.
[4]孙文静,朱文球,王业祥,等.基于卷积网络特征的逆向稀疏建模的目标跟踪[J].计算机应用与软件,2018,35(3):167-172.
[5]唐艳丽,苏燕辰.数字图像处理在距离测量中的应用[J].中国测试,2009,35(1):63-67.
[6]廖婷,蒲国林,彭小利.一种基于Radon变换及快速傅里叶变换的图像配准方法[J].计算机应用与软件,2018,35(3):182-186.
[7]徐家伟,张重阳.面向无人驾驶的高速公路指路标志字符检测[J].计算机应用与软件,2018,35(2):224-229,266.
[8]杨立娜.基于相位相关理论的最大互信息图像配准[D].西安:西安电子科技大学,2010.
[9]黄志勇,陈一民.基于频域相位相关的自适应光学图像配准算法[J].计算机应用与软件,2016,33(5):166-168.
[10]许雷,俞锋.一种基于相位相关法及数学形态学方法的眼底血管图像自动拼[J].生物医学工程学杂志,1998(3):286-290.
[11]韩峻峰,王帅.基于双目立体视觉技术的汽车测距系统实现[J].计算机应用与软件,2016,33(9):227-230.
[12]王静,张军华,单联瑜,等.相位相关速度分析方法研究及效果比较[J].物探化探计算技术,2010,32(2):164-167.
[13]王志强,程红,孙文邦,等.一种基于梯度最大值相位相关的航空影像自动配准算法[J].地理与地理信息科学,2008,24(6):111-112.
[14]陈敏,许雪林.基于Open CV的互相关图像测速方法研究[J].科学技术与工程,2010,10(15):3570-3573.
[15]王宏兴,霍玉洪.奇异值分解的教与学[J].淮南师范学院学报,2015(3):125-127.
[16]牛亚坤,玉振明,李陶深.一种近似奇异值分解的观测矩阵优化方法[J].计算机应用与软件,2016,33(1):195-197,262.
[2]Bradski G.学习Open CV[M].北京:清华大学出版社,2009.
[3]王胜辉,刘鹏,冯宏恩,等.基于Open CV的高压电气设备放电紫外成像检测图像量化参数提取[J].计算机应用与软件,2016,33(9):167-170,189.
[4]孙文静,朱文球,王业祥,等.基于卷积网络特征的逆向稀疏建模的目标跟踪[J].计算机应用与软件,2018,35(3):167-172.
[5]唐艳丽,苏燕辰.数字图像处理在距离测量中的应用[J].中国测试,2009,35(1):63-67.
[6]廖婷,蒲国林,彭小利.一种基于Radon变换及快速傅里叶变换的图像配准方法[J].计算机应用与软件,2018,35(3):182-186.
[7]徐家伟,张重阳.面向无人驾驶的高速公路指路标志字符检测[J].计算机应用与软件,2018,35(2):224-229,266.
[8]杨立娜.基于相位相关理论的最大互信息图像配准[D].西安:西安电子科技大学,2010.
[9]黄志勇,陈一民.基于频域相位相关的自适应光学图像配准算法[J].计算机应用与软件,2016,33(5):166-168.
[10]许雷,俞锋.一种基于相位相关法及数学形态学方法的眼底血管图像自动拼[J].生物医学工程学杂志,1998(3):286-290.
[11]韩峻峰,王帅.基于双目立体视觉技术的汽车测距系统实现[J].计算机应用与软件,2016,33(9):227-230.
[12]王静,张军华,单联瑜,等.相位相关速度分析方法研究及效果比较[J].物探化探计算技术,2010,32(2):164-167.
[13]王志强,程红,孙文邦,等.一种基于梯度最大值相位相关的航空影像自动配准算法[J].地理与地理信息科学,2008,24(6):111-112.
[14]陈敏,许雪林.基于Open CV的互相关图像测速方法研究[J].科学技术与工程,2010,10(15):3570-3573.
[15]王宏兴,霍玉洪.奇异值分解的教与学[J].淮南师范学院学报,2015(3):125-127.
[16]牛亚坤,玉振明,李陶深.一种近似奇异值分解的观测矩阵优化方法[J].计算机应用与软件,2016,33(1):195-197,262.