本篇論文中,我們提出一個人機互動系統,使人能夠通過攝影鏡頭與電腦進行互動並針對物件實體做影像處理以辨識其目前操作的完成進度,然後對其下一步驟所需之對應動作提示,以達到步驟教學目的。 本篇論文主要分為四個部分,第一步部分是系統架構流程圖為介紹整個系統得運作流程。第二部分是關於對目標物件的影像處理,包含了擷取目標物件影像訊息、顏色識別處理和色彩數據的比對處理。第三部分是辨識目標物件目前的完成度,其中包括物件目前完成度的辨別方式與步驟提示與顯示。最後是系統的模擬結果。 在本論文中,我們主要的貢獻有下列幾點: (1) 對目標物件做影像處理 本篇論文中提出的方法不同於其他擴增實境的應用,其中不使用ARcode來顯示對應的虛擬動畫,而是選擇直接對目標物件擷取影像與取得數據來做處理。 (2) 提供一個學習的介面 在本論文中,提出一個介面是透過攝影鏡頭與電腦進行互動,並達到能夠以一步驟一步驟提示的教學介面。
In this thesis , we propose a new scheme application to use image process combing the augmented reality that allow people to interact with computer through the webcam to learn how to assembly objects by step. This thesis can be divided into four part. First, we will introduce our algorithms to show the system. Second, is about the object image process that include the catch the objects and color identify then color data matched. Third, we introduce the recognition of the degree of completion including the tips animation and tips display. Finally, the simulations of our result will be presented. In this thesis , the results of our research are as follows: (1) Direct image process for processing the target object Our system is different with ather AR application. We don’t use the AR code technology to determine AR code to show corresponding virtual animation. In stend , use direct imaging process for processing the target object behind the screen display of the corresponding application. (2) Provide a learning interface We provide a interface through the webcam to show the tips for the to learn how to assembly object step by step.