基于视频的行人检测研究论文.pdfVIP

基于视频的行人检测研究 摘 要 随着中国经济的快速发展，公路交通日益发达，更加便利，但交通事故频繁发生，并 且有着进一步恶化的趋势，这个问题是亟待解决的。为了保证行人的人生安全，保证交通 的有序通畅，保证中国社会的和谐安定，视频的行人检测技术研发势在必行。 本论文主要是针对于基于视频的行人检测进行了研究，主要内容包括：(1)对视频文 件进行预处理，将视频一帧一帧地分离开来，以图片的形式进行保存，同时对其进行灰度 化和去噪，加强视频图像的对比度，便于之后的目标识别和提取。(2)目标图像的提取方 法主要有背景差分法和帧间差分法，这两种方法各自有自己的优点和缺点，可以通过将两 者相互结合，提取出完整的目标人物。(3)当然二值化后的目标很可能产生空洞和斑点， 致使图像走形，我们可以通过形态学处理，填充目标的空洞，从而呈现于实体相一致的二 值图像。 关键词：行人检测；背景差分法；帧间差分法；MATLAB；视频 ThePedestrianDetectionBasedVideo Abstract Withtherapid development ofChinaseconomy,qualityoflifecontinuestoimprove,road transport increasingly developed, more convenient, but accidents occur frequently, and has a furtherdeterioration,theproblemissolved.Inordertoensurethesafetyofpedestriansinlife,to ensure smooth and orderly traffic, to ensure harmony and stability of Chinese society, video pedestriandetectiontechnologydevelopmentimperative.Inaddition,insomelargesupermarkets, shoppingmalls,cinemas,especiallybanks,arelikelytoexisttheft,robberyevil,inordertocrack down on crime and help police crack cases more effectively, we need on these occasions installationofpedestriandetectionsystem. Thispaper ismainly aimed atvideo-basedpedestrian detectionhas been studied,themain contentsinclude:(1)Pre-processingofthevideofile,thevideoframeby frameseparated,inthe form of pictures to save, while its gray, strengthen the contrast of video images, object recognition and extraction to facilitate later. (2) Extraction methods mainly target image background subtraction and frame difference method, these two methods each have their own advantages and disadvantages,you can combinethe two to each other,to extract the complete target figure.(3)Ofcourse,afterthetargetbinary islikelytohavevoids andspots,causingthe imagetotake shape,wecanmorphologicalprocessing, fillingtheemptygoal,therebyrendering theentityconsistentbinaryim