文档介绍：多数据融合的车辆检测算法的研究
摘要
随着科学技术的发展和用户需求的多样化,智能系统越来越多的被人们用在日常生活和科技防御中,而智能系统中较为常见的是智能监控系统的应用,它对人们的生活产生了越来越的影响。智能监控系统顾名思义是通过视频录像,再从中检测出所需要的景物。实现智能视频监控的第一步就是从指定的监控场景中检测并提取出运动物体(比如运动的人和车辆等)。而这些运动区域的准确提取是后续的目标跟踪、识别、分类等算法顺利进行的关键前提。然而,在实际的智能视频监控系统中,由于各种光线的存在,场景中会产生大量的阴影,并且阴影与产生阴影的运动物体具有相同的运动特征,这就使得阴影和图像结合在一起,影响我们后续的图像处理,而采用单一的空间技术来检测景物时,人们大多采用目测目标的方法,这就造成了我们很难客观的找出视频中的指定景物。因此采用多数据融合的图像检测方法渐渐进入人们的事业,并且越来越受到人们的重视。这一技术已经逐渐成为该领域热点与难点。
关键词:图像处理,多数据融合,车辆检测,Matlab应用
Vehicle Detection Algorithm Based on Data fusion
ABSTRACT
Science and technology development and the diversification of user needs, more and more intelligent systems are used in everyday life people and technology in defense, andintelligen-
ce system is the mon application of intelligent monitoring system, its people's lives had more impact. Intelligent monitoring system suggests that it is through the video recording, and then detected from the scene needed. The first step in intelligent video surveillance is to monitor the scene from the specified to detect and extract moving objects (such as movement of people and vehicles, etc.). And these movements accurately extract the region is the follow- up tracking, identification, classification key prerequisite for the smooth algorithm. However, in practical intelligent video surveillance system, the existence of a variety of light, the scene will have a lot of shadows and the shadow cast a shadow of moving objects with the same movement characteristics, which makes the shadows and images together, affect our subseq- uent image processing, and the use of space technology to detect a single scene, people most- ly used method of visual targets, which resulted in difficult to find an objective specified in the video scene. Therefore, image data fusion using multiple detection methods getting into people's cause, and more and more attention. This technology has gradually e the focus and difficulty in this area.
KEY WORDS: Image processing, integration of multiple data, Ve