摘要
该文分析了数字图像处理主要研究领域中边缘检测的方法，即LoG算子、Prewitt算子和Canny算子。针对车栽摄像机复杂环境下的行人边缘检测进行研究，希望从复杂背景中有效地检测出行人轮廓边缘。首先用Matlab／Simulink作为工具，建立车载摄像机平台下的行人边缘检测模型。接着利用该模型对车载摄像机所采集到的一段视频进行试验。最后提出了一种相邻奇数帧动态背景消除的行人边缘检测方法。从实验结果可以看出我们方法的有效性。
In this paper, we analyze the edge detection method of the mare research areas ot cUgntai image processing, that LoG operator, Prewitt operator and Canny operator. We focus on pedestrian edge detection of complex environment for car camera, hoping to effectively detect pedestrian edge from complex environment. First, we establish the pedestrian edge detection model on vehicle camera platform with the Matlab/Simulink as a tool. Then we test a video capture from the on vehicle camera using the model. In the end, we propose a pedestrian edge detection method base on a background of Adjacent odd frames dynamic background erasing. Experimental results show our method high oerformance.
作者
许焕明
高向东
XU Huan-ming, GAO Xiang-dong（1.Guangdong Communication Polytechnic, Guangzhou 510800, China ; 2.The College of Mechanical ＆ Electronic Engineering of Guangdong Unive~ity of Technology, Guangzhou 510090, China）
出处
《电脑知识与技术》
2011年第12期8919-8922,共4页
Computer Knowledge and Technology
