摘要: 针对实时场景中运动目标检测的问题,提出一种改进算法。结合背景差分与帧间差分判定图像是否存在运动目标的过程中,引入形态腐蚀算子抑制大量孤立小噪声,采用基于矩形块的目标判定方法提高检测准确度,通过实时背景更新提高自适应性。实验结果表明改进算法可靠性高、抗干扰性强、背景适应性强。
关键词:
运动目标检测,
背景差分,
帧间差分,
背景更新,
矩形块判定
Abstract: This paper proposes an improved algorithm aiming at moving object detecting in real-time scene. In the process of combing background differencing with interframe difference to dual decide moving object in the image being or not, it introduces morphological erosion operator to restrain a lot of isolated noise, uses the rectangular decision method to improve the precision of detection. Real-time update can also realize to enhance the adaptability of background. Experimental result shows that improved algorithm has high reliability, strong anti-interference and background adaptability.
Key words:
moving object detection,
background difference,
interframe difference,
background update,
rectangular decision
中图分类号:
王明吉;肖桂丹. 抗干扰的运动目标检测算法[J]. 计算机工程, 2010, 36(3): 232-233,.
WANG Ming-ji; XIAO Gui-dan. Anti-interference Moving Object Detection Algorithm[J]. Computer Engineering, 2010, 36(3): 232-233,.