摘要
为了提高运动目标检测的准确性和效率,提出了基于改进的三帧差分法检测运动目标的算法。该算法对传统的三帧差分法进行了改进,结合了单高斯模型背景提取、背景自适应更新和自适应阈值提取的方法,从而解决了传统三帧差分法中可能出现的无法检测出完整的运动目标的问题。背景的自适应更新和自适应阈值的提取减少了光线亮度变换以及噪声对运动目标检测带来的影响。实验表明,该算法与传统的三帧差分法相比可以更加完整地检测出运动目标,并且可以有效地避免出现"漏检"等情况,提高了运动目标检测的效率和准确性。
To accurately and efficiently improve the moving objects detection, this paper proposes an improved three frame difference algorithm, which improves the traditional three frame difference and combines background extraction and self-adaptive threshold extraction. This algorithm solves the problem that can＇t detect the entirety of the moving objects in the traditional three frame difference. Self-adaptive background update and self-adaptive threshold extraction reduce the effect of the sunshine and noise. Experimental results demonstrate that the algorithm can detect the entirety of the moving objects and avoid undetected error. This algorithm can accurately and efficiently detect the moving objects detection.
出处
《电视技术》
北大核心
2013年第1期151-153,共3页
Video Engineering
基金
国家"973"计划项目(2005CB321901)
软件开发环境国家重点实验室开放课题(BUAA-SKLSDE-09KF-03)
关键词
三帧差分法
背景自适应更新
自适应阈值
运动目标检测
three frame difference
self-adaptive background update
self-adaptive threshold
moving objects detection
