毕业论文

打赏
当前位置: 毕业论文 > 计算机论文 >

基于openCV运动物体检测的方法研究

时间:2018-03-23 16:53来源:毕业论文
通过对图像的处理,检测出目标,并进行定位。植入opencv图像处理库,在分析了背景差分,帧差分,光流场法等多种检测方法特点的基础上,运用这几种不同的方法分别对运动目标进行

摘要:随着科技的发展,作为人类替代品的机器人在越来越多的领域中得到了广泛的应用,机器人技术也日益发达。机器人视觉是机器人系统中非常重要的组成部分,它主要包括了图像的采集,预处理和图像检测几部分。本文主要研究了机器人视觉系统中对图像的处理和对运动物体的检测。本文对颜色模型和视频图像序列进行了研究,通过对图像的调整和变化,使图像更容易识别和处理,防止了因各种干扰对图像参数的不利影响,以利于信息选取或抑制。运用visual studio软件,通过对图像的处理,检测出目标,并进行定位。植入opencv图像处理库,在分析了背景差分,帧差分,光流场法等多种检测方法特点的基础上,运用这几种不同的方法分别对运动目标进行检测处理,并实验证明了其可行性和可靠性。20048
关键词:灰度化,预处理,模糊处理,帧差法,高斯背景建模,光流
 Study on the method of object detection based on openCV
Abstract: With the development of science and technology,as the human alternative, the robots have been widely used in more and more fields. And the robot technology is also increasingly developed.Robot vision is a very important part in a robot system. It mainly includes image acquisition, pre-processing and image detection part. This paper mainly studies the robot vision system for image preprocess and detection of moving objects. This paper studied the color model and the video image sequence, the image adjustments and changes, make the image more easily identify and deal with, avoid the interference of various adverse effects on the image parameter, in order to facilitate the information selection or inhibition. The use of Visual Studio software, through image processing, target detection, and position。Then we implanted OpenCV image processing library, in the analysis of the background difference, frame difference,optical flow method based features and other detection methods, the moving target detection are using thesedifferent methods, and the experiments proved the feasibility and reliability.

源¥自%六^^维*论-文+网=www.lwfree.cn


key words:Gray, Pretreatment,Fuzzy processing,Frame difference method,The background difference method,Optical flow method

 目录
1. 绪论    5
 1.1 研究背景与意义    5
 1.2 研究现状分析    5
 1.3研究目的及论文组织    6
2.视频图像的处理基础    8
 2.1色彩基础    8
  2.2颜色空间及其模型    9
  2.2.1 RGB颜色模型    10
  2.2.2 HSI颜色模型    11
  2.2.3RGB与HIS颜色模型的转化    11
 2.3图像灰度化    13
 2.4视频图像序列    14
 2.5本章总结    14
3. 基于OPenCV的图像预处理    16
 3.1 平滑处理    16
  3.1.1 简单模糊    17
  3.1.2中值模糊    17
  3.1.3高斯模糊    18
  3.1.4双边滤波    19
  3.1.5  4种模糊处理的选择    19
 3.2图像二值化处理    20
 3.3图像形态学    20
 3.4本章总结    22
4.运动目标的提取与检测    23
 4.1帧差法    23
 4.2背景差分    25
  4.2.1 高斯背景模型的建立    26
  4.2.2  opencv中高斯模型的应用    27
 4.3光流场法    28
  4.3.1 Lucas-Kanade方法    29 基于openCV运动物体检测的方法研究:http://www.lwfree.cn/jisuanjilunwen/20180323/11678.html
------分隔线----------------------------
推荐内容