Digital Signal Processing and Artificial Intelligence for Automatic Learning Hung-Yuan Chung, Yao-Liang Chung, Chih-Yen Liang 1 2 2022 https://doi.org/10.6025/dspaial/2022/1/2/93-99 https://www.dline.info/dspai/fulltext/v1n2/v1n2_5.pdf In this study, a novel system for correcting posture was developed through the application of wearable sensing technology. This system comprises three subsystems, namely, a smart necklace, smart phone, and notebook computer. The smart necklace houses an MPU- 6050 sensor that can collect and analyze gravitational acceleration data when the necklace is worn, and use that data to determine a user’s upper body posture. When the necklace detects poor posture, it will send a reminder to the user’s smart phone, which is able to receive such messages from the necklace via a mobile app that was developed in this study. This app can also be used to set parameter values such as the standard values (base values for posture assessment) of the necklace and the timing of reminders, and to activate the posture image calibration function on the notebook computer. The notebook computer is used to read the relevant data via a depth camera, identify the user’s skeletal structure and joint reference points, and perform reference point computations, after which signals will be sent to the necklace in order to calibrate the necklace’s standard values. This proposed system allows for: 1) the self-correction of posture to be carried out without the use of corrective clothing that is heavy, thick, and stuffy; 2) the quick calibration of the necklace’s standard values through the use of posture imaging; and 3) the elimination of complex wiring and the effective application of the Internet of Things through the implementation of wireless communication between the interfaces of the smart necklace, smart phone, and notebook computer