智慧型手持裝置與生活應用的結合是近年來流行的議題,本論文主要嘗試提供一種智慧型手持裝置架構的實作,以照相與網路結合智慧型手持裝置運算功能,建置一個行動車牌辨識系統。系統主要採Open Source的Android平台裝置,及Java應用程式為基礎核心來建置雲端主機。手機部份提供UI畫面、方便的人機互動選取車牌、以及影像粗定位功能。雲端主機部份實現了OCR圖片處理、代登網站查詢、和資料紀錄功能。藉由雲端主機預先處理,僅將必要資訊回傳手機,以加快網路傳輸。此架構模式能有效降低機器的開發、管理、與維護升級成本,實現應用服務快速整合與資源有效利用的目標。 The integration of portable devices and our everyday life is a popular topic in recent years. This thesis attempts to provide an architectural implementation of portable devices, using photographs and networks to build a mobile license plate recognition systems. The system adopted the open source code of the Android platform device and Java applications in building a cloud server. The portable device provides user interface (UI), easy selection of license plates through interaction, and rough positioning the plates via image processing algorithms. The cloud server provides OCR (optical character recognition) of license plate images, web site inquiry, and data recording. By the preprocessing of the cloud server, only necessary information will be sent back to the portable device in order to speed up network transmission. This architectural model is able to effectively reduce the cost of development, management, and maintenance. Furthermore, it also achieves the goal of rapid integration and effective utilization of resources.