计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (14): 164-165.

• 图形、图像、模式识别 • 上一篇    下一篇

帧差累积和减背景相结合的运动对象分割方法

韩亚伟,张有志,李庆涛,熊春彬   

  1. 山东大学 信息科学与工程学院,济南 250100
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-05-11 发布日期:2011-05-11

Moving object segmentation method using accumulated frame difference and background subtraction

HAN Yawei,ZHANG Youzhi,LI Qingtao,XIONG Chunbin   

  1. School of Information Science and Engineering,Shandong University,Jinan 250100,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-05-11 Published:2011-05-11

摘要: 针对低比特率的多媒体视频序列,提出了一种综合利用帧差累积和背景减法来进行运动对象分割的方法。由一种改进的帧差累积方法得到初步的运动对象区域,通过背景减法得到运动对象区域,把由两种方法得到的运动对象区域相结合取得完整准确的结果,二值化后再经过形态学处理和二次扫描填充即可得到运动对象掩模,用原图像的灰度值填充该区域。实验表明,该方法快速,准确,并有一定的应用价值。

关键词: 帧差累积, 运动对象区域, 背景减法, 形态学处理, 运动对象掩模

Abstract: A moving object segmentation method is presented for low bit rate multimedia video sequences.It utilizes accumulated frame difference and background subtraction.The initial moving object region is gotten by an improved accumulated frame difference method,then another moving object region is obtained by the background subtraction method.The two moving object regions are combined to obtain complete and accurate result.After binarization and morphological processing,the moving object mask can be obtained by twice scanning and filling.The moving object region is filled with original values.Experiments show that this method is fast and accurate and has application value.

Key words: accumulated frame difference, moving object region, background subtraction, morphological processing, moving object mask