Human motion and posture detection based on the multi-nodes sensors will help people to realize the law of body's movement through obtaining the kinetic parameters, so that it can be used as a quantitative analysis tool for some specific applications such as health care, sport training and military affairs, etc. Compared with optical method, inertial sensors for body motion analysis have advantages of non-limitation of space and lower cost. This paper describes an inertial sensors-based human motion monitoring system, and the human body communication (HBC) was designed to transmit the data among the nodes. According to the algorithm of quaternion for data fusion with the data of the inertial sensor, the three-dimensional angles of the human arm could be calculated. Then the data would be transmitted to the computer to reconstruct the posture of human's arm. In the system, HBC method had been used, which utilizes the human body as the communication channel to transmit the kinetic data for posture reconstructure. Consequently, the circuit of IBC device based on capacitive coupling had been designed in the paper. The result shows that the inertial sensors could obtain the three-dimensional attitude angle of human's arm, and the IBC device successfully received the digital base data at the bit rate of 115.2 kbps. The vivo experiments were implemented to verify the feasibility of the system.