运动图像分析毕业论文

《运动图像分析毕业论文》由会员分享，可在线阅读，更多相关《运动图像分析毕业论文（39页珍藏版）》请在毕设资料网上搜索。

1、 摘摘 要要 运动目标的检测是计算机视觉和图像编码领域的重要研究之一， 无论是在军 事，还是医学和科研等领域都有广泛的应用。 本文主要介绍了一些目前较为常用运动目标检测的方法， 并以数字图像处理 技术为基础，设计出高效的运动目标检测方法。目前，运动目标检测主要有以下 三种方法：基于帧间差分法，基于背景差分的方法和基于光流场的方法。本文根 据所处理的视频图像的特点以及可行性，主要研究帧间差分法和背景差法。以帧 间差分法为例，利用相邻帧图像间强相关性的特点，采用帧间差分目标检测的方 法。在对图像差分之前，先进行图像预处理，再进行相邻帧差分得到运动信息， 再对目标的边缘轮廓和背景进行数学形态学滤波来

2、去除背景噪声， 提取出目标轮 廓，该方法的优点是操作简便、速度快、易于硬件实现。在论文后期引入了人脸 的定位、人脸的识别功能，其中脸部的定位主要根据脸部肤色的亮度值、脸部形 状特点、 眼球亮度值以及眼部对称性等特点， 利用最小矩阵查找脸部区域并标记； 而人脸的识别则根据输入图片与训练库中的图片做比照，完成图片的识别。 关键词：图像处理；目标检测；帧间差分：阈值分割；人脸识别 Abstract Moving target detection is one of the important research field of computer vision and image coding, wh

3、ether it is in the military, medicine and research and other fields have a wide range of applications. This paper describes some of the more commonly used moving target detection method based on digital image processing technology, the design of efficient moving target detection method. , Moving tar

4、get detection are mainly the following three methods: based on inter-frame difference method based on background difference method and the method based on optical flow field. Based on the characteristics and the feasibility of the processed video image mainly study the inter-frame difference method

5、and the background difference method. Inter-frame difference method, for example, the using adjacent frame image is a strong correlation between the characteristics of the target detection frame difference method. In before the image difference, the first image pre-processing, then the adjacent fram

6、es, the differential motion information, morphological filtering to remove background noise to extract the target contour on the edge contour of the target and background, the advantages of this method is simple, fast, and easy hardware implementation. Introduced late in the thesis of the positioning of the face, the face recognition function, wherein the positioning of the face is mainly based on the luminance value of facial skin, the facial shape characteristics of the lumi