近年來由於Microsoft 公司的Kinect 體感偵測器藉由人體動作的偵測與辨識而廣泛應用於體 感遊戲。在商業產品調查過程中經常必須要觀察消費者對產品反應的行為以作為廠商對產品改進 的重要參考依據，本論文的主要研究動機即在於結合Kinect 體感偵測器與個人電腦來設計一個 可以即時偵測與判斷消費者臉部表情的辨識系統，如此可以節省相關的人力及物力資源。我們的 辨識系統是藉由Kinect 體感偵測器來即時偵測受測者臉部動態特徵點以及嘴巴、眉毛、眼睛等 並運用Kinect 的軟體開發工具所來設計一個軟體程式以辨識消費者臉部的表情，目前可以判斷 的表情有高興、驚訝、悲傷、與自然表情等四種，依照我們的實驗結果可以得知消費者臉部表情 的平均辨識度高達87.5%。

In recent years, due to the Microsoft's Kinect motion detectors, the motion detection games have been widely used. It is very important for product improvement by observing the behavior of the consumer reaction to the product. The main motivation of this paper is to develop a Kinect motion detector based real-time facial expression detection and recognition system, thus saving the human effort to observe the consumers. So far, by using our tools and the Kinect motion detector, the happy, surprise, sadness, and natural facial expressions of the consumers can be judged in real-time. By our experiments, the recognition rate of the facial expressions is up to 87.5% in average.