Development of an Image Processing-based iOS Interactive Mora Application

[[abstract]]近年來配備多項感測元件與技術之智慧型個人行動裝置 (如智慧型手機與平板電腦等) 已逐漸普及，從「易用性」的角度研發對使用者最友善的人機介面應用，例如適切地整合運用手勢辨識、體態感測、自然語言等互動多媒體處理技術，達到簡單、易學、直覺操作與高效率的目標，已逐漸成為一種 App 開發趨勢。本論文研究以開發一套行動裝置上的互動式猜拳遊戲為目標，首先發展一種基於手型輪廓的數種特徵配合影像處理程序以辨識出手指數目之演算法，當中運用了使用者的膚色資訊與多種影像前處理技術，再利用行動裝置上的各項裝置如攝像鏡頭、多點觸控螢幕與高性能處理器等進行實作。以在行動裝置上能即時實現並有良好辨識率為前提，所提出的手指數目辨識演算法這項核心技術除了克服了複雜背景、攝像遠近與手勢旋轉等問題，更巧妙地精簡運算複雜度以提升執行效率，這些效能已由 iPhone 實機應用獲得驗證。[[abstract]]Recently intelligent mobile devices with multiple sensors and technologies such as smartphones and tablet PCs are spread widely and gradually From the perspective of “usability” to study and develop user-friendly human-machine interface application has become the trend of App development; for example the use of interactive multimedia processing technology such as integrating gesture-recognition application motion sensing and natural language achieves the goal of easy learning simple and intuitive operation and high efficiency The aim of the current study is to develop an interactive Mora game for mobile devices Based on the hand contours with various features coordinated with the image processing program such as the use of information on skin color and of image pre-processing techniques an algorithm of recognizing the number of fingers is developed first and then implemented actually by utilizing equipment like camera multi-touch screen and high-performance processor Under the premise of instant realization and well recognition rate the core technology proposed algorithm for recognizing finger numbers which has been applied to and verified by iPhone can not only overcome the problems like complicated background variable camera distance and gesture rotation but also ingeniously simplify the computational complexity to increase implementation efficienc