Vehicle Detection by Sparse Deformable Template Models

Vehicle detection is an important problem in computer vision. Several applications including robotics, surveillance and automotive safety are related to vehicle detection. In this paper, we build up a vehicle detection system by combing the active basis model and logistics regression. Active basis model provides a robust and reasonable representation for cars, while logistic regression gives us an efficient classifier for big data. A detailed system framework is presented and some experiments show good performance in both accuracy and speed of the developed system.