基于ARM9的嵌入式图像处理平台的设计及应用

发布时间：2018-05-26 15:04

本文选题：嵌入式Linux操作系统 + 数字图像处理　；参考：《南京理工大学》2012年硕士论文

【摘要】：本文的主要目的是搭建一个较为通用的嵌入式数字图像处理平台,主要解决了嵌入式数字图像处理系统的选型、搭建、使用等问题。本文首先分析了嵌入式图像处理系统的最新发展,结合实际需求,分析本次硬件平台所采用的ARM920T核心板资源,继而以模块化的模式介绍本文所设计的接口板,并给出搭建完成后硬件平台的结果。其次,本文分析了选择Linux操作系统作为本次开发所用嵌入式操作系统的原因,详细说明了在ARM9内核上安装嵌入式Linux操作系统的步骤；根据遇到的实际情况,搭建基于嵌入式Linux操作系统的网络文件系统,并分为服务器端和用户端详细介绍搭建的步骤,极大的减低了开发的难度。第三,本文针对平台通用性的特点,向嵌入式平台移植了数字图像处理开源库OpenCV2.0以及多平台图形界面设计工具Qtopia2.2.0,文章详细介绍两个工具的移植步骤和使用方法,描述了使用中容易出现的错误并给出解决方法,为嵌入式平台下图像复杂算法的实现和实时显示提供了一条途径。第四,本文在搭建的嵌入式数字图像处理平台上实现了基于经典算法的条形码识别功能和人脸检测功能,在文章中给出实现的结果。最后,本文总结了研究中取得的成果,并阐述了本文不足之处,提出了本课题的进一步研究方向。
[Abstract]:The main purpose of this paper is to build a general embedded digital image processing platform, mainly to solve the embedded digital image processing system selection, construction, use and other problems. This paper first analyzes the latest development of embedded image processing system, combining with the actual demand, analyzes the ARM920T core board resources used in this hardware platform, and then introduces the interface board designed in this paper in modular mode. The results of the hardware platform are given. Secondly, this paper analyzes the reasons why the Linux operating system is chosen as the embedded operating system used in this development, and explains the steps of installing the embedded Linux operating system on the ARM9 kernel in detail. Build the network file system based on the embedded Linux operating system, and divided into the server and the user side to introduce the steps of building in detail, greatly reduce the difficulty of development. Thirdly, according to the general characteristics of the platform, this paper transplanted the open source digital image processing library OpenCV2.0 and the multi-platform graphical interface design tool Qtopia 2.2.0to the embedded platform. The paper introduces the porting steps and using methods of the two tools in detail. This paper describes the errors that are easy to occur in use and gives a solution, which provides a way for the realization and real-time display of complex image algorithms on embedded platform. Fourthly, this paper realizes the function of bar code recognition and face detection based on classical algorithm on the embedded digital image processing platform. The results are given in this paper. Finally, this paper summarizes the achievements of the research, expounds the shortcomings of this paper, and puts forward the further research direction of this subject.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP368.1;TP391.41

【参考文献】