摘要
针对经典视觉背景提取算法因初始帧存在运动目标易产生鬼影以及对扰动背景适应性差的问题,提出一种改进ViBe算法;利用改进三帧差分法和最小外接矩形定位初始帧运动目标,并通过局部初始化的方法进行鬼影抑制;在背景模型初始化阶段,定义灰度相似函数从时域和空域信息中中等比例选取像素点建立背景模型,增强背景模型的鲁棒性;在前景检测检测阶段,通过平均差法衡量样本集合的离散度,构建自适应分割阈值代替原有的固定分割阈值以适应背景扰动;实验表明,改进算法可以有效抑制鬼影产生并且提高算法在扰动背景下的适应性和检测准确度。
Aiming at the problem that the classical visual background extraction(ViBe)algorithm is prone to ghost image due to the presence of moving object in the initial frame and poor adaptability of dynamic background,an improved ViBe algorithm is proposed.Using the improved three-frame difference method and minimum circumscribed rectangle,the initial frame moving target is located,and the ghost image is suppressed by the method of local initialization;In the initial stage of the background model,a grayscale similarity function is defined to select the pixels in a moderate proportion from the temporal and spatial information and establish the background model to enhance the robustness of the background model;In the foreground detection stage,the average difference method is used to measure the dispersion of the sample set,and an adaptive segmentation threshold is constructed to replace the original fixed segmentation threshold and adapt to the background disturbance.Experiments show that the improved algorithm can effectively suppress the ghost and improve the performance in the disturbance adaptability and detection accuracy.
作者
陈从平
江高勇
张力
凌阳
郁春明
闫焕章
张屹
CHEN Congping;JIANG Gaoyong;ZHANG Li;LING Yang;YU Chunming;YAN Huanzhang;ZHANG Yi(School of Mechanical and Rail Transportation Changzhou University,Changzhou 213164,China;School of Materials Science&Engineering Changzhou University,Changzhou 213164,China)
出处
《计算机测量与控制》
2022年第12期105-111,共7页
Computer Measurement &Control
基金
江苏省产业前瞻与关键核心技术-碳达峰碳中和科技创新专项资金项目(BE2022044)
国家自然科学基金(51875053)。
关键词
运动目标检测
ViBe算法
帧差法
鬼影
自适应阈值
moving target detection
frame difference
ViBe algorithm
ghost
adaptive threshold
作者简介
陈从平(1976-),男,湖北荆州人,博士研究生,教授,博士生导师,主要从事机器视觉、3D打印方向的研究;张屹(1976-),男,甘肃兰州人,博士研究生,教授,博士生导师,主要从事机电系统现代设计方法、机电传动与控制系统设计方向的研究;通讯作者:江高勇(1996-),男,江苏宿迁人,硕士研究生,主要从事机器视觉方向的研究。
