基于帧差法和背景差法融合的车流量检测方法  

Traffic Flow Detection Method Based on Fusion of Frames Differencing and Background Differencing

作　　者：许伦辉[1,2] 卜文萍[2] 陈衍平[2] 黄艳国[1,2] 

机构地区：[1]华南理工大学,广东广州510640 [2]江西理工大学电气工程与自动化学院,江西赣州341000

出　　处：《计算机仿真》2012年第2期353-357,共5页Computer Simulation

摘　　要：研究优化车流量检测准确度问题。针对运动目标速度和外界环境都是影响车流量检测准确性,容易造成车流量的漏检和误检等。为了克服传统算法所存在的缺陷,在现有算法的基础上,提出了一种融合帧差法和背景差法的智能车流量检测方法。首先利用帧间差分方法为主,结合减背景方法为辅,然后通过一种迭代阈值分割法滤除噪声并对背景进行实时更新。完成了多车道的车流量检测,并进行了仿真,结果得到多组数据,并提高了计算准确率。仿真结果表明,改进方法可有效地提高了车流量检测精度。Research the problem of optimizing flow detection accuracy. The main factors affecting the flow detection are moving target speed and external environment, which is likely to cause the miss and false detection of traffic volume. In order to overcome the shortcomings of traditional algorithm, the paper proposed a fusion frame difference and background subtraction method for intelligent traffic detection on the basis of the existing algorithm. The mainly method is frame difference method, combined with method of reducing background. First, the background noises were filtered out and the background was updated in real time through an iterative thresholding method. A multi - lane traffic flow detection was completed to get multiple sets of data and calculation accuracy. The method is effective for improving the traffic flow measurement accuracy according to the simulation results, and is simple and feasible.

关 键 词：车流量 背景差法 帧差法 目标检测 

分 类 号：TB24[一般工业技术—工程设计测绘]