基于改进的几何约束算法与卷积神经网络的车辆检测
|
摘要
论文提出了一种基于改进的几何约束算法有效结合卷积神经网络(CNN)的车辆检测方法。首先对几何约束算法进行改进,避免该算法重复的矩阵运算,从而进一步提高该算法效率。根据改进的几何约束算法计算出车辆的感兴趣区域,然后在该区域内提取Haar-like特征,通过Adaboost分类器初步检测得到候选框。之后用训练好的卷积神经网络模型对目标候选框进行分类。实验结果表明,该方法能够有效地减少车辆检测时间,提高车辆检测的精度,并且对多种光照条件,部分遮挡,姿态变化等具有一定的鲁棒性。
In this paper,a vehicle detection method based on improved geometric constraint algorithm combined with convolution neural network is proposed.Firstly,the geometric constraint algorithm is improved to avoid the matrix operation,in order to further improve the efficiency of the algorithm.According to the improved geometric constraint algorithm,the region of interest of the vehicle is obtained.And the Haar-like feature is extracted in the region,then the Adaboost classifier is used for the preliminary detection.The convolution neural network model is trained by the vehicle training set.The convolution neural network model is used to classify candidate regions.The experimental results show that this method can reduce the vehicle detection time and improve the accuracy of vehicle detection effectively,and it is robust for a variety of lighting conditions,partial occlusion and posture.
In this paper,a vehicle detection method based on improved geometric constraint algorithm combined with convolution neural network is proposed.Firstly,the geometric constraint algorithm is improved to avoid the matrix operation,in order to further improve the efficiency of the algorithm.According to the improved geometric constraint algorithm,the region of interest of the vehicle is obtained.And the Haar-like feature is extracted in the region,then the Adaboost classifier is used for the preliminary detection.The convolution neural network model is trained by the vehicle training set.The convolution neural network model is used to classify candidate regions.The experimental results show that this method can reduce the vehicle detection time and improve the accuracy of vehicle detection effectively,and it is robust for a variety of lighting conditions,partial occlusion and posture.
引文
[1]王亮.光流技术及其在运动目标检测和跟踪中的应用研究[D].长沙:国防科学技术大学,2007.WANG Liang. Optical Flow Technology and Its Applica-tion in Motion Target Detection and Tracking[D]. Chang-sha:National University of Defense Technology,2017.
[2]Arróspide J,Salgado L,Nieto M,et al. On-board robustvehicle detection and tracking using adaptive quality eval-uation[J]. 2008:2008-2011.
[3]Hoffmann C. Fusing multiple 2D visual features for vehi-cle detection[C]//Intelligent Vehicles Symposium. IEEE,2006:406-411.
[4] Rezaei M,Terauchi M,Klette R. Robust Vehicle Detec-tion and Distance Estimation Under Challenging LightingConditions[J]. IEEE Transactions on Intelligent Transpor-tation Systems,2015,16(5):2723-2743.
[5]文学志,方巍,郑钰辉.一种基于类Haar特征和改进AdaBoost分类器的车辆识别算法[J].电子学报,2011,39(5):1121-1126.WEN Xuezhi,FANG Wei,ZHENG Yuhui. A Vehicle Rec-ognition Algorithm Based on Haar-like Feature and Im-proved AdaBoost Classifier[J]. Journal of Electronics,2011,39(5):1121-1126.
[6] Dalal N,Triggs B. Triggs,B. Histograms of Oriented Gra-dients for Human Detection. In:CVPR[J]. 2005,1(12):886-893.
[7] Viola P,Jones M. Rapid object detection using a boostedcascade of simple features[C]//Computer Vision and Pat-tern Recognition,2001. CVPR 2001. Proceedings of the2001 IEEE Computer Society Conference on. IEEE,2003:I-511-I-518 vol.1.
[8]Liu C,Wechsler H. Gabor feature based classification us-ing the enhanced fisher linear discriminant model for facerecognition[M]. IEEE Press,2002.
[9]Sivaraman S,Trivedi M M. Looking at Vehicles on theRoad:A Survey of Vision-Based Vehicle Detection,Tracking,and Behavior Analysis[J]. IEEE Transactionson Intelligent Transportation Systems,2013,14(4):1773-1795.
[10]Cheon M,Lee W,Yoon C,et al. Vision-Based VehicleDetection System With Consideration of the Detecting Lo-cation[J]. Intelligent Transportation Systems IEEETransactions on,2012,13(3):1243-1252.
[11]Zakaria Z,Suandi S A. Face detection using combinationof Neural Network and Adaboost[C]//TENCON 2011-2011 IEEE Region 10 Conference. IEEE, 2011:335-338.
[12]Cortes C,Vapnik V. Support-vector networks[C]//Ma-chine Learning. 1995:273-297.
[13]Viola P,Jones M J. Robust Real-Time Object Detection[C]//International Workshop on Statistical and Computa-tional Theories of Vision--Modeling,Learning,Com-puting,and Sampling. 2001:87.
[14]Krizhevsky A,Sutskever I,Hinton G E. ImageNet classification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems. Curran Associates Inc. 2012:1097-1105.
[15] Sudowe P,Leibe B. Efficient Use of Geometric Constraints for Sliding-Window Object Detection in Video[C]//International Conference on Computer Vision Systems. Springer-Verlag,2011:11-20.
[16]Jones M,Viola P. Fast multi-view face detection[J].Proc of Computer Vision&Pattern Recognition,2003,11(1):276-286.
[17]Lienhart R,Maydt J. An extended set of Haar-like features for rapid object detection[C]//International Conference on Image Processing. 2002. Proceedings. IEEE,2002:I-900-I-903 vol.1.
[18]Freund Y,Schapire R E. A decision-theoretic generalization of on-line learning and an application to boosting[C]//European Conference on Computational Learning Theory. Springer Berlin Heidelberg,1995:23-37.
[19]Bengio Y,Lecun Y. Convolutional Networks for Images,Speech,and Time-Series[J]. 1995.
[20]Jia Y,Shelhamer E,Donahue J,et al. Caffe:Convolutional Architecture for Fast Feature Embedding[J].2014:675-678.
[2]Arróspide J,Salgado L,Nieto M,et al. On-board robustvehicle detection and tracking using adaptive quality eval-uation[J]. 2008:2008-2011.
[3]Hoffmann C. Fusing multiple 2D visual features for vehi-cle detection[C]//Intelligent Vehicles Symposium. IEEE,2006:406-411.
[4] Rezaei M,Terauchi M,Klette R. Robust Vehicle Detec-tion and Distance Estimation Under Challenging LightingConditions[J]. IEEE Transactions on Intelligent Transpor-tation Systems,2015,16(5):2723-2743.
[5]文学志,方巍,郑钰辉.一种基于类Haar特征和改进AdaBoost分类器的车辆识别算法[J].电子学报,2011,39(5):1121-1126.WEN Xuezhi,FANG Wei,ZHENG Yuhui. A Vehicle Rec-ognition Algorithm Based on Haar-like Feature and Im-proved AdaBoost Classifier[J]. Journal of Electronics,2011,39(5):1121-1126.
[6] Dalal N,Triggs B. Triggs,B. Histograms of Oriented Gra-dients for Human Detection. In:CVPR[J]. 2005,1(12):886-893.
[7] Viola P,Jones M. Rapid object detection using a boostedcascade of simple features[C]//Computer Vision and Pat-tern Recognition,2001. CVPR 2001. Proceedings of the2001 IEEE Computer Society Conference on. IEEE,2003:I-511-I-518 vol.1.
[8]Liu C,Wechsler H. Gabor feature based classification us-ing the enhanced fisher linear discriminant model for facerecognition[M]. IEEE Press,2002.
[9]Sivaraman S,Trivedi M M. Looking at Vehicles on theRoad:A Survey of Vision-Based Vehicle Detection,Tracking,and Behavior Analysis[J]. IEEE Transactionson Intelligent Transportation Systems,2013,14(4):1773-1795.
[10]Cheon M,Lee W,Yoon C,et al. Vision-Based VehicleDetection System With Consideration of the Detecting Lo-cation[J]. Intelligent Transportation Systems IEEETransactions on,2012,13(3):1243-1252.
[11]Zakaria Z,Suandi S A. Face detection using combinationof Neural Network and Adaboost[C]//TENCON 2011-2011 IEEE Region 10 Conference. IEEE, 2011:335-338.
[12]Cortes C,Vapnik V. Support-vector networks[C]//Ma-chine Learning. 1995:273-297.
[13]Viola P,Jones M J. Robust Real-Time Object Detection[C]//International Workshop on Statistical and Computa-tional Theories of Vision--Modeling,Learning,Com-puting,and Sampling. 2001:87.
[14]Krizhevsky A,Sutskever I,Hinton G E. ImageNet classification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems. Curran Associates Inc. 2012:1097-1105.
[15] Sudowe P,Leibe B. Efficient Use of Geometric Constraints for Sliding-Window Object Detection in Video[C]//International Conference on Computer Vision Systems. Springer-Verlag,2011:11-20.
[16]Jones M,Viola P. Fast multi-view face detection[J].Proc of Computer Vision&Pattern Recognition,2003,11(1):276-286.
[17]Lienhart R,Maydt J. An extended set of Haar-like features for rapid object detection[C]//International Conference on Image Processing. 2002. Proceedings. IEEE,2002:I-900-I-903 vol.1.
[18]Freund Y,Schapire R E. A decision-theoretic generalization of on-line learning and an application to boosting[C]//European Conference on Computational Learning Theory. Springer Berlin Heidelberg,1995:23-37.
[19]Bengio Y,Lecun Y. Convolutional Networks for Images,Speech,and Time-Series[J]. 1995.
[20]Jia Y,Shelhamer E,Donahue J,et al. Caffe:Convolutional Architecture for Fast Feature Embedding[J].2014:675-678.