基于最小误差阈值自适应的ViBe改进算法  

An improved ViBe algorithm based on adaptive minimum error threshold

作　　者：张磊 孟亮[1] ZHANG Lei;MENG Liang(School of Computer Science and Technology,Taiyuan University of Technology,Taiyuan 030024,China)

机构地区：[1]太原理工大学计算机科学与技术学院,山西太原030024

出　　处：《现代电子技术》2019年第4期161-165,共5页Modern Electronics Technique

基　　金：山西省自然科学基金应用基础研究项目(201701D121057)~~

摘　　要：针对ViBe算法无法快速抑制鬼影现象的问题,提出一种基于最小误差阈值自适应的ViBe改进算法。文中通过计算每一帧图像的最小误差分割阈值来对鬼影像素点进行判别并消除,加快了鬼影的抑制速度。并且根据样本方差的大小动态调整样本距离判定阈值,提高了该算法检测复杂场景时的准确率。此外,通过计算像素点的梯度值有效地优化了背景模型的更新速率。实验结果表明,改进后的ViBe算法有效地提高了鬼影的消除速度,并且算法的准确率也得到了提升。Since the ViBe algorithm cannot rapidly suppress the ghost phenomenon,an improved ViBe algorithm based on the adaptive minimum error threshold is proposed.The ghost pixels are discriminated and eliminated by computing the minimum error segmentation threshold of each image frame,so as to accelerate the ghost suppression speed.The threshold for determining sample distances is dynamically adjusted according to the variance values of samples,so as to improve the complex scene detec-tion accuracy of the algorithm.The update rate of the background model is effectively optimized by calculating the gradient values of pixels.The experimental results show that the improved ViBe algorithm can effectively improve the speed of ghost elimination,and the accuracy of the algorithm is also improved.

关 键 词：智能监控 运动目标检测 背景建模 ViBe算法 最小误差阈值分割 鬼影消除 像素梯度 

分 类 号：TN911.73-34[电子电信—通信与信息系统]