基于帧间差分的背景提取与更新算法

摘要

In view of the distortional background extracted by time-averaged method, an improved algorithm that combined with the frame-difference method is presented in this paper. During background updating, this paper uses multi-level background update strategy, it can avoid the effect of new objects to the background and eliminate the ghosts caused by the move-away of background objects or by the move-in of foreground objects. Experimental result shows that the proposed algorithm can be used to extract backgrounds effectively and quickly when the scene changes, and it has better effect.%当运动目标较多时,时间平均法提取的背景存在模糊和拖影现象.为此,提出一种基于帧间差分的背景提取与更新算法.该算法采用基于像素级和帧级的多级自适应背景更新策略,能够克服光线缓慢变化和剧烈变化对背景的影响,并及时消除由于物体移入和移除产生的鬼影.实验结果表明,该算法能快速有效地对背景进行更新,提取的背景图像效果较好.