先验信息光流法在运动目标检测中的应用  

Based on Optical Flow Method with Priori Information in the Application of Detecting Moving Target Tracking

作　　者：李倩倩[1] 刘彦隆[1] 

机构地区：[1]太原理工大学通信与嵌入式实验室,太原030024

出　　处：《火力与指挥控制》2015年第10期156-160,共5页Fire Control & Command Control

摘　　要：针对运动目标的智能跟踪检测提出一种基于先验信息的光流检测法。首先,通过利用区域生长法获取的道路先验信息并结合帧差法确定运动目标的大致范围。其次,采用光流法对处理后的图像进行分割,并由形态学方法对进行图像处理后检测出运动目标。最后,基于MATLAB平台对其进行仿真计算。结果表明,该方法克服了传统光流法在运动目标检测中易受干扰的问题,并为军事人员对运动目标跟踪检测提供了有效的辅助。The paper proposes an improved method that based on optical flow algorithm in moving target tracking detection application. Firstly,it obtain the road priori information which use region growing method and range of moving target which through frame difference algorithm. Secondly,target moving objects is detected that through optical flow algorithm and morphological processing. Finally,through MATLAB platform simulation reveals that the new method overcomes interference problem in moving target detection that exist traditional optical flow method. Meanwhile it is provides the effective auxiliary for military.

关 键 词：运动目标智能跟踪检测 光流法 区域生长 先验信息 MATLAB 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]