A Fast Video Vehicle Detection Approach Based on Improved Adaboost Classifier

Aiming at the problem that traditional vehicle detection methods often fail to ensure the accuracy and speed simultaneously, a vehicle detection method based on background difference and improved Gentle Adaboost classifier is proposed. Firstly, the foreground region is obtained by using the background difference method, and the morphological processing is applied properly to get better candidate foreground regions. Then, the cascaded Adaboost classifiers are used to detect multi-scale vehicles in these regions. In this paper, we adopt effective search strategy, which can greatly reduce the number of search windows, and further improve the detection speed. The experimental results show that the proposed method not only can obtain high accuracy, but also has strong real-time performance. Precisely, the highest accuracy reaches to 96.0% and the highest detection speed reaches to 51.4 FPS.

[1]  Meng Joo Er,et al.  A Novel Approach for Vehicle Detection Using an AND–OR-Graph-Based Multiscale Model , 2015, IEEE Transactions on Intelligent Transportation Systems.

[2]  J. Friedman.  Special Invited Paper-Additive logistic regression: A statistical view of boosting , 2000 .

[3]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[4]  Qingxiang Wu,et al.  Moving vehicle detection based on optical flow estimation of edge , 2015, 2015 11th International Conference on Natural Computation (ICNC).

[5]  Marek Wojcikowski,et al.  FPGA-Based Real-Time Implementation of Detection Algorithm for Automatic Traffic Surveillance Sensor Network , 2012, J. Signal Process. Syst..

[6]  Byung-Ryong Lee,et al.  Front-view car detection and counting with occlusion in dense traffic flow , 2015 .

[7]  Oihana Otaegui,et al.  Adaptive Multicue Background Subtraction for Robust Vehicle Counting and Classification , 2012, IEEE Transactions on Intelligent Transportation Systems.

[8]  Shuguang Li,et al.  Video-Based Traffic Data Collection System for Multiple Vehicle Types , 2012 .

[9]  Qiuxia Wu,et al.  Real-time vehicle detection with foreground-based cascade classifier , 2016, IET Image Process..

[10]  Weixing Wang,et al.  An accurate vehicle counting approach based on block background modeling and updating , 2014, 2014 7th International Congress on Image and Signal Processing.