The invention relates to a multi-feature fusion attitude recognition system and method. The system includes a management terminal, a Cloud Server, a Wireless Networksand Human Body Nodes, wherein thehuman body node comprises a thorax node, a food node L, and a foot node R; The foot node L comprises a second single chip microcomputer, a second 2.4 G module, a second power supply module and a firstforce-sensitive sensor group; and the foot node R comprises a third single chip microcomputer, a second pressure sensor, a third 2.4 G module, a third power supply module and a second force-sensitivesensor group. In accordance with that present invention, Attitude angle and height difference percentage can detect the posture change of upper body and foot, monitor the change of human body centerof gravity combined with plantar pressure characteristics, and identify posture by using parameters obtained from cloud computing training, which can effectively identify the daily behavior posture ofthe body, and can be queried on the terminal, and has a wide application prospect.