基于运动目标尺寸检测的实时图像处理与显示  

Real-time Image Processing and Display in Size of Objects Detection Based on Moving Objects

作　　者：翟亚宇 潘晋孝[1] 刘宾[1] 陈平[1,2] 

机构地区：[1]中北大学电子测试技术国家重点实验室,太原030051 [2]中国科学院自动化研究所中国科学院分子影像重点实验,北京100190

出　　处：《计算机测量与控制》2014年第11期3794-3797,共4页Computer Measurement &Control

基　　金：国家自然基金(61171179;61227003;61301259);山西省自然科学基金(2012021011-2);高等学校博士学科点专项科研基金资助课题(20121420110006);山西省回国留学人员科研资助项目(2013-083);山西省高等学校优秀创新团队支持计划资助

摘　　要：运动目标尺寸实时检测与显示是运动图像检测和应用视觉研究领域的一个重要课题,而图像处理则是其核心内容;文章针对采集到的图像进行一系列的处理,最终实时显示图像并获得运动目标的尺寸等信息;首先针对采集图像时电子设备的不稳定性和检测模块的不一致性,进行中值滤波和增益校正同时将处理图像实时显示出来;然后对其整体图像进行分割与二值化、角点检测和边缘拟合处理,获得目标的尺寸信息;最后利用VC++将整体处理图像及尺寸信息显示出来,以便于图像理解和模式识别;实验证明,文章提出的算法、软件的设计很好地滤除不相关的目标,保存感兴趣目标,具有良好的实时性、精确性,能够满足当代工业检测设备中的要求。The real-time detection and display in the size of moving objects is an important topic in applied vision and moving image detection,and the image processing is one of the core algorithms of size detection and display.This paper mainly focuses on the processing and display of the target dynamic images that are collected in real-time,and obtained a real-time image processing result of moving objects.We arrange this article as follows：Firstly,electronic equipment is instability during image acquisition process and detection module is not inconsistent,so we will employ the median filtering and gain calibration to the real-time image of the moving objects to make the image closer to the real object,meanwhile the real-time image is displayed.Secondly,image segmentation,image binarization,corner detection and edge fitting will be used to the whole image,then the size of object is computed.Finally,VC＋＋ will be used to display the processed image and size information.Experimental results indicates that the proposed algorithm and the software design can filter uninteresting moving objects in vedio timely and meanwhile save the interesting objects,simultaneously,it has better real-time performance and accuracy which will meet the industrial testing equipments＇ needs.

关 键 词：运动目标尺寸检测 实时图像处理与显示 增益校正 角点检测 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置] TP391.4[自动化与计算机技术—控制科学与工程]