Vehicle detection is a foundational and significant task in video surveillance systems. In this paper, a vehicle detection method using a deformable model and symmetry is proposed. First, we learn the active basis model (ABM) from the target training sample set by using the shared sketch algorithm. Then, we utilize the edge information obtained by ABM and HSV color information to do symmetry analysis for vehicle objects. The ABM can detect vehicles in various poses, shapes, and sizes for its deformability. By doing edge and color symmetry analysis, subtle difference between two images and environment noises can be adapted. The results of experiments indicate that our approach is capable of detection different vehicles and localization vehicle in bad environment. What's important, the detection results support the capability of the proposed method to enable the introduction of novel intelligent transportation systems applications.
Vehicle detection method based on active basis model and symmetry in ITS
2013-10-01
695885 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Multipart Vehicle Detection Using Symmetry-Derived Analysis and Active Learning
Online Contents | 2015
|Multipart Vehicle Detection Using Symmetry-Derived Analysis and Active Learning
Online Contents | 2016
|Symmetry-based monocular vehicle detection system
British Library Online Contents | 2012
|