透過手勢辨識之互動人機介面

An Interactive Human-Computer Interface Through Hand Gesture Recognition

李政信(Cheng-Hsin Li)

指導教授 ： 黃銘智

國立臺北科技大學/設計學院/創新設計研究所/碩士(2013年)

https://doi.org/10.6841/NTUT.2013.00416

摘要

隨著辨識技術的進化，人與機械之間的互動變得更加豐富而多元化，自然人機介面令人機互動領域增添了不少想像與願景。人與電腦的直接溝通儼然形成一種趨勢，擁有控制意義的手勢辨識則成為自然互動中最重要的領域。
　　手勢辨識的研究範圍很廣，本文從自然互動的觀念出發，以深度影像分析為切入點。透過深度影像之特性，探討手部追蹤與即時手勢辨識之概念。本研究對深度攝影機所讀取之深度影像進行分析運算，建立一個可追蹤使用者指尖的辨識系統，並以此系統驗證一到五的數字手勢辨識。
　　本文隨後於案例驗證中，將此手勢辨識系統應用於互動遊戲與電腦控制，證明手勢辨識於互動人機介面應用之機會與價值。

關鍵字

手勢辨識 ； 自然互動 ； 深度影像

並列摘要

With the evolution of recognition technology, interaction between human and machine is becoming more abundant and diversified. Definition of natural user interface was added more imagination and vision to the interactive field. Human communication with the computer directly becomes a trend. Hand gesture recognition means controlling significance could be the most important area of natural interaction.
　　There are many researches of hand gesture recognition, this paper aims to implement the concept of natural interaction. The depth image is an entry point to hand gesture recognition. We explore the technologies of hand tracking and fingertips recognition for real-time interaction through the features of depth image. In this research, we analyze and compute the depth image from the depth camera. We build a hand gesture recognition system which can be tracking the motion of fingers and recognizing the number of fingers on two of user’s hands. We use this recognition system to demonstration the numbers gesture from one to five fingers.
　　Then we apply this system to an interactive game and computer remote control, so that the effect and the value of an interactive human computer interface through hand gesture recognition are proved in our demonstrated examples.

並列關鍵字

Hand Gesture Recognition ； Natural Interaction ； Depth Image

參考文獻

[2]呂緯昱(2010)，「具有現場膚色偵測以指尖與夾角為基礎的手勢辨識方法」，台北科技大學資訊工程系，碩士論文，台北，2010

[3]黃堯彬、黃詰琳、詹寶珠(2008)，「使用溫感影像做靜態手勢辨識」，成功大學電腦與通訊工程研究所，2008年數位生活科技研討會。

[4]Argyros A. A. & Lourakis M. I. A. (2004). Real time Tracking of Multiple Skin-Colored Objects with a Possibly Moving Camera, in proc. ECCV’04, Springer-Verlag, vol. 3, pp. 368-379.

[5]Argyros A. A. & Lourakis M. I. A. (2006). Binocular Hand Tracking and Reconstruction Based on 2D Shape Matching, 18th International Conference on Pattern Recognition, pp. 207-210.

[6]Barber C. B., Dobkin D. P., & Huhdanpaa H. (1996). The Quickhull Algorithm for Convex Hulls, ACM Transactions on Mathematical Software, 22(4), pp. 469-483.