摘要
在智能交通监控系统中,阴影检测是其中的重点和难点现有的方法有些是简单的以某种情况作为前提,而不能适合范围更广的情况;还有些则过于复杂,且难以计算。本文提出了一种基于FPGA的多信息相融合的运动阴影检测算法,充分发挥了FPGA的并行性,降低了传统计算机图像处理的延迟,通过多种检测信息弥补单一检测算法的不足,并提高检测精度。
Shadow detection is key and difficulty in intelligent traffic monitoring system. Some of the existing methods are premise in some cases simply, so it can not be suitable for a wider range of circumstances; Some of methods are overly complex and difficult to calculate. This paper proposes a moving shadow detection algorithm of multi-information fusion based on FPGA, it take full advantage of the parallelism of FPGA and reduce the delay of a conventional computer image processing, moreover, it compensate for the deficiencies of the single detection algorithm by a variety of detection information to improve detection accuracy.
