计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (5): 192-193.

• 图形、图像、模式识别 • 上一篇    下一篇

一种快速核密度估计背景建模方法

乔俊锋,朱 虹,史 静,孟凡星   

  1. 西安理工大学 自动化与信息工程学院,西安 710048
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2012-02-11 发布日期:2012-02-11

Fast kernel density estimation method for background modeling

QIAO Junfeng, ZHU Hong, SHI Jing, MENG Fanxing   

  1. College of Automation and Information Engineering, Xi’an University of Technology, Xi’an 710048, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-11 Published:2012-02-11

摘要: 针对核密度估计背景建模方法运算量大难以实时应用的问题,提出了一种基于背景直方图分布的快速核密度估计背景建模方法。选用三角核函数进行核密度估计,根据三角核带宽函数的截断效应,引入背景分布的直方图完成快速背景建模,在保证目标检测准确性的同时提高运算速度。测试实验结果验证了算法能够满足监控系统的实时性要求。

关键词: 背景建模, 核密度估计, 背景样本集直方图, 实时性

Abstract: For kernel density estimation is difficult to satisfy real-time applications because of large amount of calculation, this paper proposes a fast kernel density estimation method of background modeling based on the background histogram. Triangle kernel function is used to estimate the kernel density. According to the triangular truncation effect of kernel-bandwidth function, background samples histogram is built to complete the fast background modeling. The accuracy of target detection is ensured while processing speed is increased. Experimental results prove that the algorithm satisfies the real-time requirements of surveillance systems.

Key words: background modeling, kernel density estimation, histogram of the background sample set, real-timing