Arm9-based Embedded Image Processing Platform Design And Application

This thesis aims at developing general embedded system for image processing, majorly solving the problem of how to choose, construct and use such embedded system.In the thesis, a review is first made on the latest development of embedded system for image processing, and ARM920T used as kernel in the hardware platform are then introduced, followed by a module-by-module introduction to interface panels designed in the thesis, with the layout of completed hardware platform displayed.Secondly, this thesis analyzes the advantage of using Linux as the operating system(OS) for embedded system and illustrates in detail the procedure of installing Linux on ARM9. Network file system(NFS) for Linux OS is also constructed for practical use, and detailed steps in constucting NFS for user terminal and server terminal are both specified, reducing greatly the diffcultity in the development of embedded system.Thirdly, considering the broad applicability of the platform, to which two software programs are migrated, OpenCV2.0 and Qtopia2.2.0, the former being the open source library for image processing and the other being multi-platform graphical interface design tool. This thesis gives a detailed account of how to migrate and use both toos, and offers solutions to errors likely to encounter, presenting an approach of implementing sophisticated image processing algorithm in embedded system and display the results in real time.Besides, typical programs for identifying bar codes and recognizing human face are tested in the contructed embedded system for digital image processing with the results shown.At last, a conclusion is drawn on the controbution and shortcomings of this thesis, and puts forward further research efforts for this topic.