自适应的智能监控背景提取技术.docxVIP

PAGE  第  PAGE 1 页 共  NUMPAGES 24 页 PAGE \* MERGEFORMAT0 分类号：××× U D C：******-×××-(20***) ××××-0 密 级：公 开 编 号：XXX ***********大学 学位论文 自适应的智能监控背景提取技术 论文作者姓名：申请学位专业：申请学位类别：指导教师姓名（职称）：论文提交日期：  自适应的智能监控背景提取技术 摘要 智能化已是当今监控系统的主要发展方向，能否自动检测出运动目标与提取背景图像是检验一个智能监控系统的重要标志。在实际应用中，由于光线变化、物体背影等外部条件的影响会使运动物体的提取与背景图像的提取出现一定的变化，智能监控系统无法提取出理想的背景图像。本文分别运用统计中值法与多帧平均法对一段监控视频图像进行背景图像提取，运用统计学的中间值或者平均值的方式，对多张视频图像中相同像素点灰度值的变化进行计算得出该点的背景值，那么可以区分并得出该点是运动目标或是背景图像，提取出背景图像，并进一步分析处理背景图像。 关键字：智能监控；运动目标检测；背景提取；统计中值法；多帧平均法  Adaptive intelligent monitoring technology background extraction Abstract Intelligence is the main direction of development of todays surveillance systems can automatically detect moving targets and extract the background image is a test of intelligent monitoring system is an important symbol. In practice, due to changes in light, objects, and other external conditions will back to extraction with a background image of a moving object appears some changes, intelligent monitoring system can not be extracted ideal background image. In this paper, respectively, using the statistical method and the value of the multi-frame averaging period of surveillance video image background image extraction, the use of statistical median or mean way to change multiple video images in the same pixel gray values were calculated the background of this point, we can distinguish between that point and draw a background image or a moving target, thereby extracting a background image, and further analysis to monitor the background image. Key words：Intelligent monitoring；Moving object detection；Background extraction；Statistical median method；Multi-frame average method 目录 论文总页数:20页  TOC \o 1-3 \h \z \u  HYPERLINK \l _Toc451888232 1引言  PAGEREF _Toc451888232 \h 1  HYPERLINK \l _Toc451888233 1.1背景与意义  PAGEREF _Toc451888233 \h 1  HYPERLINK \l _Toc451888234 1.2国内外研究现状  PAGEREF _Toc451888234 \h 1  HYPERLINK \l _Toc451888235 2视频背景