MATLAB视频中运动目标的提取方法研究
视频中的目标检测是在人机交互领域极其重要的一个环节，目的即是将所选择目标从视频背景之中提取，并加以计算其轨迹，从而得出相关结论。在目标提取阶段，有几个因素可能会对这一过程产生影响，如：云雾、光照方向、目标的拖影等，以至于视频中运动目标的提取难度加大。
对于以上情况，大多数研究者在运动物体检测以及运动物体追踪领域主要应对方法有两种：一是对视频进行处理得到视频中所标记的目标的连续图像；二是利用人工智能神经网络对图像目标进行直接的追踪。二者相比，前者较易被大众接受且便于实现，能够高效得到目标轨迹。现在较为普遍的方法有背景减法、邻帧差法和光流法。25502
关键词 目标检测 目标提取 背景相减法
毕业论文设计说明书外文摘要
Title Video moving target extraction
Abstract
Nowadays,visual information is full of many aspects of life,the visual information has became more and more important than other information channels. Moving target tracking and detection can be achieved based on MATLAB. Get information into artificial intelligence has become an important research direction.Low Accuracy visual information would have a significant inconvenience in daily life.Therefore,install video moving target acquisition system in the key locations is especially important.
Video target detection in the field of human-computer interaction is an extremely important aspect, the purpose of it is tell out the selected target from among video background extraction,and calculate its trajectory,then comes out conclusions. In the object extraction stage, there are several factors that could have an impact on this process, such as: clouds, light direction, goals smear, etc. so that the difficulty of extracting the video moving target increases.
For the above, there are two ways that most researchers in moving object detection and tracking moving objects use in the research: First, the video is processed to obtain a moving image; the second is the use of artificial neural network image directly target tracking. Compared the two ways, the former is easier to be accepted by the public and easy to implement, and can efficiently obtain the target trajectory. Now,common methods are background subtraction, adjacent frame difference and optical flow.
Keywords Target Detection Object Extraction Background subtraction
目 次
摘要I
AbstractII
1 绪论 1
1.2 课题的研究背景及意义 1
1.3 国内外研究概况…2
1.3.1 背景差法…2
1.3.2 邻帧差法2
1.3.3 光流法…3
1.4 论文的研究内容及安排4
2 视频中运动目标追踪研究5
2.1 引言5
2.2 静态背景下运动目标的侦测算法研究 5
2.2.1 经典边缘提取检测5
2.2.2 微分算子6
2.2.3 Canny算子…7
2.3 运动目标分割7
2.3.1 图像差分法8
2.3.2 背景差分法9
2.4 Kalman滤波器9
2.5 本章小结10
3 运动目标跟踪方法研究 11
3.1 搜索算法的介绍 11
3.1.1 绝对平衡搜索法（ABS）… 11
3.1.2 归一化互相关匹配算法（NC）… 12
3.1.3 均值偏移算法原理…12
3.2 运动目标方向性判断 12
上一篇：基于扫频光源的人体血流成像研究