A Two-stage Background Reconstruction Algorithm for the Detection of Moving Targets in Traffic Scene

摘要

Background reconstruction algorithm is the basis of background subtraction algorithm which is widely used in the detection of moving targets that is one of the key issues in the Intelligent Transportation System (ITS in short). So in order to detect moving targets more completely and accurately, a robust and effective background algorithm is very important and is the urgent need. Taking into account the complex environment in traffic scene, this paper presents a two-stage background reconstruction algorithm (including background initialization and background updating) based on the study of the existing background reconstruction algorithms, and verifies that the effectiveness of the algorithm is very good by inferring from the theory and analyzing the experimental results.