设计了一种基于双目立体视觉技术的汽车车身尺寸测量系统。首先,在相机标定和校正环节利用张氏标定法和Bouguet算法分别进行了相机模型参数求解以及图像校正,消除因图像畸变而造成后续测量误差的影响;其次,结合背景差分算法与OTSU大津算法,对目标实现了较高精度的分割及轮廓提取;最后,利用寻找最小外接矩形框,结合使用SGBM立体匹配算法将匹配后得到的车体二维数据转换为三维数据,完成车辆尺寸的测量。实验结果表明,本系统的整体测量误差在6%以内,可以有效完成车身尺寸测量。
Abstract
In the paper,a vehicle body dimension measurement system based on binocular stereo vision technology is designed.Firstly,in the camera calibration and correction link,the camera model parameter solution and image correction are carried out using Zhang's calibration method and Bouguet algorithm,respectively,to eliminate the influence of subsequent measurement errors caused by image distortion.Secondly,combined with the background difference algorithm and the OTSU algorithm,high-precision segmentation and contour extraction are achieved for the target.Finally,the vehicle size measurement is completed by finding the smallest bounding rectangle and using the SGBM stereo matching algorithm to convert the two-dimensional data of the car body obtained after matching into three-dimensional data.The experiment results show that the overall measurement error of this system is within 6%,and the measurement of body size can be effectively completed.
关键词
双目视觉 /
车身尺寸测量 /
立体校正 /
SGBM立体匹配
Key words
binocular vision /
vehicle dimension measurement /
stereo correction /
SGBM stereo matching
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 刘琦.道路交通安全管理研究[D].青岛:中国海洋大学, 2012.
[2] Hoegg T,Lefloch D,Kolb A.Time-of-Flight camera based 3D point cloud reconstruction of a car[J].Computers in Industry,2013, 64(9):1099-1114.
[3] 邢丽,张弘治,陈曦,等.基于双目视觉的薄壁零件变形量测量[J].光子学报,2020,49(7):14.
[4] 董鹏,周烽,赵悰悰,等.基于双目视觉的水下海参尺寸自动测量方法[J].计算机工程与应用,2021,57(8):8.
[5] Fang J,Xu S,Yang Y,et al.Localization and measurement method of continuous casting slab model based on binocular vision[J].Microwave and Optical Technology Letters,2019,62(12).
[6] 孙郑芬,吴韶波.机器视觉技术在工业智能化生产中的应用[J].物联网技术,2020,10(8):4.
[7] Zhang Z.A Flexible New Technique for Camera Calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1334.
[8] Fusiello A,Trucco E,Verri A.A compact algorithm for rectification of stereo pairs[J].Machine Vision & Applications,2000,12(1):16-22.
[9] 刘波,王凌,金以慧.差分进化算法研究进展[J].控制与决策,2007,22(7):9.
[10] Otsu N.A Threshold Selection Method from Gray-Level Histograms[J].IEEE Transactions on Systems Man & Cybernetics,2007,9(1):62-66.
[11] 白明,庄严,王伟.双目立体匹配算法的研究与进展[J].控制与决策,2008,23(7):9.
[12] 王昕,马岩,杨剑,等. 区域立体匹配算法的实现及改进[J].光学精密工程,2008(10):2002-2008.
[13] H Hirschmüller.Stereo Processing by Semiglobal Matching and Mutual Information[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(2):328-341.
[14] Kim J,Kolmogorov V,Zabih R.Visual correspondence using energy minimization and mutual information[C]//Ninth IEEE International Conference on.IEEE,2003.
[15] Birchfield S, Tomasi C.A Pixel Dissimilarity Measure That Is Insensitive to Image Sampling[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,1998,20(4):401-406.
基金
*中国—黑山科技合作委员会第3届例会交流项目(No.3-7);长沙理工大学“双一流”科学研究国际合作拓展项目(No.2019ic18);柔性电子材料基因工程湖南省重点实验室开放基金(No.202005)。
PDF(3639 KB)
