智能交通监控系统中的自适应背景更新算法研究

Self-Adaptive Background Update Algorithm in an Intelligent Traffic Monitoring System

在基于视觉的交通监控系统中,场景图像的分割即背景与前景目标的分割是车辆检测中的第一步,在整个处理中占有很重要的地位,它的准确性直接关系到后继工作的难易程度,甚至整个系统的成败.该文提出一种自适应背景更新算法,此算法能更简单、更鲁棒地实现分割.

In visionbased traffic monitoring system, segmentation of background and objects is the first step in vehicle detection. Its accuracy directly affects the complexity of the subsequent processes, and even determines the effectiveness of the whole system. This paper presents an algorithm of selfadaptive background update for traffic images in visionbased monitoring systems. The algorithm provides a simple and robust solution to the segmentation problem.