本論文主要利用穿戴式裝置，進行駕駛行為姿態感測資訊蒐集與校正，校正後的感測資料將作為駕駛行為姿態建模之輸入參數，再透過隱馬可夫模型(Hidden Markov Model；HMM)進行駕駛行為姿態建模，可正確辨識出7種駕駛者頭部姿態變化。其中建模環境同時考量靜態室內環境與動態車內環境，並透過實車環境測試，估算模型辨識效能，同時根據測試結果進行模型參數調校，以提升駕駛行為姿態辨識效果。最後利用所建立之駕駛行為辨識模型，將駕駛姿態辨識結果與車內智慧型裝置相關功能進行匹配，以提供駕駛更便利之車內操控服務。

The purpose of this study is to present an approach to analyze and recognize the driver posture. By using the wearable device, the sensing data of the driving behavior postures can be collected and calibrated. The calibrated data is as the input parameters of driving behavior modeling. Then, the Hidden Markov Model (HMM) is utilized to establish the driving behavior posture models. The HMM model can recognize 7 kinds of head postures. Moreover, the static indoor environment and dynamic in-car environment are considered in modeling process. The recognition performance is evaluated by field trial, and these testing results are utilized to calibrate and adjust to enhance the recognition performance. Finally, the constructed driving behavior model can apply to in-car service to provide driver more convenient control service.