- 學位論文
基於深度與色彩資訊之三維自然使用者介面設計與晶片實現
Design and Chip Implementation of Depth and Color-based Three Dimensional Natural User Interface
摘要
隨著近幾年來觸控螢幕與浮空手勢操作的發展，人們逐漸由傳統的鍵盤與滑鼠操作介面，逐漸轉變成為更符合人類的直覺操作方式。浮空的手部操作介面利用彩色與深度攝影機擷取空間中的影像，利用所擷入的彩色資訊與深度資訊進行手部的追蹤與手勢的辨識，藉由分析手部當前的狀況來進行電腦與使用者之間的互動。 本論文使用結合深度與彩色攝影機進行空間中物體的捕捉，利用所擷入的彩色資訊與深度資訊進行影像的分析，過程包含了手部的切割、手部偵測、手部的追蹤、手指偵測與動作的辨識，經由所設計的演算法流程得知當前手部的位置、手部的移動方向等資訊，系統除了可以判斷較大面積的使用者手勢辨識，並更進一步的分析手指當前的狀態與關節點的位置，進行精細的手指控制，並設計了一系列的使用者互動介面，如手指的點擊、手勢手指圖像的放大縮小、圖形的旋轉等操作介面來進行使用者的互動操控，其手指點擊平均準確率為94.6%，手部放大縮小與旋轉手勢平均準確率為98.2%，互動功能的應用可以運用在相當多不同的地方，例如、電腦的操控、電視的轉台控制、電子廣告刊板的浮空手部互動…等。 此外，我們也透過數位積體電路設計的流程，設計硬體系統的架構，經由對演算法的分析與最佳化實現其硬體，藉由攝影機擷取影像與晶片的處理，使影像處理速度更加流暢。
並列摘要
Accompany with touch screen panel and mid-air control have been developed in recent years, people gradually change their usage from tradition keyboard and mouse to the intuitive manner. Mid-air hands operation interface uses RGB-D camera to capture images from space. Then it uses captured color and depth information to track hands and gesture to interact with computer. In this thesis, we used RGB-D based camera to capture objects in space and analyzed the data with color and depth information. The series of processing includes hand segmentation, hand detection, hand tracking, finger detection and action recognition. The system can find current hands location and hand movement information and so on. Furthermore we can use finger to control objects by analyze the current status of finger and determine the finger articulation. We also design a series of interactive interface like finger clicking, picture zooming/rotation by gesture, picture zooming by finger and so on to operate with computer. The accuracy of finger clicking and picture zooming/rotation by gesture are 94.6% and 98% respectively. The system can be applied to computer control, channel change of TV, and interact with advertisement board, etc. Besides, we designed system hardware by analyzing the algorithm and optimizing its architecture. We implemented a chip via the “Cell-based IC Design Flow.” By using chip to process images that captured by camera, the whole system will perform real time processing.