计算机技术与发展
計算機技術與髮展
계산궤기술여발전
COMPUTER TECHNOLOGY AND DEVELOPMENT
2015年
1期
37-40
,共4页
目标检测%背景差法%图像处理%欧氏距离%背景建模
目標檢測%揹景差法%圖像處理%歐氏距離%揹景建模
목표검측%배경차법%도상처리%구씨거리%배경건모
object detection%background subtraction%image processing%Euclidean distance%background modeling
为了从包含大量冗余信息的监控视频中快速查找到运动目标,提出了一种改进的背景差值目标检测算法。首先,通过灰度化和中值滤波对视频图像进行预处理;其次,对视频帧进行抽样统计,计算各个对应像素点的灰度值的中值,建立背景模型;再次,通过大量的实验确定合适的阈值后,计算当前帧与背景模型之间欧氏距离的相对差值,并由此判断前景帧和背景帧;最后,将含有运动目标的图像或视频截取出来。实验结果表明,该方法可以更加准确有效地检测目标,可用于视频监控(如生活小区、铁路交通、仓库的监控视频等)中的目标检测。
為瞭從包含大量冗餘信息的鑑控視頻中快速查找到運動目標,提齣瞭一種改進的揹景差值目標檢測算法。首先,通過灰度化和中值濾波對視頻圖像進行預處理;其次,對視頻幀進行抽樣統計,計算各箇對應像素點的灰度值的中值,建立揹景模型;再次,通過大量的實驗確定閤適的閾值後,計算噹前幀與揹景模型之間歐氏距離的相對差值,併由此判斷前景幀和揹景幀;最後,將含有運動目標的圖像或視頻截取齣來。實驗結果錶明,該方法可以更加準確有效地檢測目標,可用于視頻鑑控(如生活小區、鐵路交通、倉庫的鑑控視頻等)中的目標檢測。
위료종포함대량용여신식적감공시빈중쾌속사조도운동목표,제출료일충개진적배경차치목표검측산법。수선,통과회도화화중치려파대시빈도상진행예처리;기차,대시빈정진행추양통계,계산각개대응상소점적회도치적중치,건립배경모형;재차,통과대량적실험학정합괄적역치후,계산당전정여배경모형지간구씨거리적상대차치,병유차판단전경정화배경정;최후,장함유운동목표적도상혹시빈절취출래。실험결과표명,해방법가이경가준학유효지검측목표,가용우시빈감공(여생활소구、철로교통、창고적감공시빈등)중적목표검측。
In order to find the moving object from the surveillance video which contains a lot of redundant information quickly,an im-proved object detection algorithm based on background subtraction is put forward. Firstly,video images are preprocessed by graying and median filtering. Secondly,by statistical sampling of video frames,the median value of gradation value of each corresponding pixel point is calculated,and then a background model can be established. Thirdly,this algorithm can determine the foreground frame and the back-ground frame through the difference of the relative Euclidean distance between the current frame and the background model,but before that,an appropriate threshold must be determined by a large number of experiments. Finally,the images or videos containing moving ob-jects will be clipped out. Experimental results show that this method can detect objects more effectively and accurately,and can be used for object detection ( such as residential,rail transport,warehouse surveillance video,etc) in the video surveillance.