摘要
提出一种基于深度神经网络的摄像机标定方法,实现了复杂环境下平面区域内的灵活、高精度标定。无需进行数据特征提取或分类,仅通过优化网络结构、超参数与训练算法,深度神经网络便能得到快速有效的训练。实验结果表明,相较于张正友标定法与浅层神经网络标定法,该方法在大范围、多拍摄角度和高畸变条件下均能达到更高的标定精度,镜头存在高畸变时,633mm×763mm标定范围内的平均标定误差仅为0.1471mm。
This study proposes a new deep neural network based camera calibration method that achieves flexible,high-precision calibration in complex environments,without having to classify or extract features from input data.By optimizing the network structure,hyperparameters,and training algorithms,the deep neural network can be quickly and effectively trained.The experimental results confirm that,compared with Zhang′s calibration method and the shallow neural network,the proposed method can achieve high calibration accuracy under a wide range of imaging conditions involving multiple shooting angles or high distortion.For the images produced using a highly distorted lens,the proposed method achieves an average calibration error of only 0.1471 mm over the calibration range of 633 mm×763 mm.
引文
[1]Zhang Z Y.Camera calibration[M]∥Ikeuchi K,Computer Vision:A Reference Guide.Boston:Springer,2014:76-77
[2]Tsai R.A versatile camera calibration technique for high-accuracy 3D machine vision metrology using offthe-shelf TV cameras and lenses[J].IEEE Journal on Robotics and Automation,1987,3(4):323-344.
[3]Zhang Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1334.
[4]Zou P P,Zhang Z L,Wang P,et al.Binocular camera calibration based on collinear vector and plane homography[J].Acta Optica Sinica,2017,37(11):1115006.邹朋朋,张滋黎,王平,等.基于共线向量与平面单应性的双目相机标定方法[J].光学学报,2017,37(11):1115006.
[5]Wu Q H,Chen H,Zhu S S,et al.Fast calibration method of large dimension measurement system based on multi-camera array[J].Acta Optica Sinica,2018,38(12):1215002.吴庆华,陈慧,朱思斯,等.一种多相机阵列大尺寸测量系统快速标定方法[J].光学学报,2018,38(12):1215002.
[6]Zhu F,Yu F S,Wu Y M,et al.Analysis of attitude calibration precision of P4Pcamera[J].Acta Optica Sinica,2018,38(11):1115005.朱帆,于芳苏,吴易明,等.P4P法相机姿态标定精度分析[J].光学学报,2018,38(11):1115005.
[7]Ramalingam S,Sturm P.A unifying model for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(7):1309-1319.
[8]Hornik K,Stinchcombe M,White H.Multilayer feedforward networks are universal approximators[J].Neural Networks,1989,2(5):359-366.
[9]Wen J,Schweitzer G.Hybrid calibration of CCDcameras using artificial neural nets[C]∥1991IEEEInternational Joint Conference on Neural Networks,November 18-21,1991,Singapore.New York:IEEE,1991:337-342.
[10]Jun J,Kim C.Robust camera calibration using neural network[C]∥Proceedings of IEEE.IEEE Region 10Conference.TENCON 99.′Multimedia Technology for Asia-Pacific Information Infrastructure′(Cat.No.99CH37030),September 15-17,1999,Cheju Island,Korea.New York:IEEE,1999:694-697.
[11]Memon Q,Khan S.Camera calibration and threedimensional world reconstruction of stereo-vision using neural networks[J].International Journal of Systems Science,2001,32(9):1155-1159.
[12]LüC H,Zhang Z Y,An P.Camera calibration for stereo vision based on neural network[J].Chinese Journal of Mechanical Engineering,2003,39(9):93-96.吕朝辉,张兆杨,安平.基于神经网络的立体视觉摄像机标定[J].机械工程学报,2003,39(9):93-96.
[13]Tian Z,Zhang Q,Xiong J L,et al.Large-scale camera calibration with neural network[J].Acta Optica Sinica,2011,31(4):0415001.田震,张玘,熊九龙,等.基于神经网络的大范围空间标定技术[J].光学学报,2011,31(4):0415001.
[14]Liu X J,Li X J,Wang W Y,et al.Large-scale binocular camera calibration combining neural network with virtual target[J].Optical Technique,2017,43(4):314-318.刘小娟,李学军,王文韫,等.基于神经网络的虚拟靶标大视场双目相机标定技术[J].光学技术,2017,43(4):314-318.
[15]LeCun Y,Bengio Y,Hinton G.Deep learning[J].Nature,2015,521(7553):436-444.
[16]Hinton G,Deng L,Yu D,et al.Deep neural networks for acoustic modeling in speech recognition:the shared views of four research groups[J].IEEESignal Processing Magazine,2012,29(6):82-97.
[17]Hinton G E,Osindero S,Teh Y W.A fast learning algorithm for deep belief nets[J].Neural Computation,2006,18(7):1527-1554.
[18]Glorot X,Bordes A,Bengio Y.Deep sparse rectifier neural networks[C]∥Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics,April 11-13,2011,Fort Lauderdale,FL,USA.Cambridge:PMLR,2011,15:315-323.
[19]Bergstra J,Bengio Y.Random search for hyperparameter optimization[J].Journal of Machine Learning Research,2012,13(1):281-305.