Human Action Recognition Based on Inertial Sensors and Complexity Classification

Related Work

Z.Zhang et al (Zhang Z et al., 2009, Zhang Z et al., 2011) raised a hierarchical information fusion algorithm, which can estimate the human gestures. In the analysis, the algorithm has geometric constraints on human bone, and they introduced particle filter algorithm in the information fusion process. Although this method recognition rate is higher, its calculation is relatively complicated. Especially for the embedded systems with few resources and low computing, its implementation is difficult. Furthermore, Hao Yang (Yang et al., 2011) and Luinge (Luinge et al., 2007) gained the gesture relationship between the carrier sensing coordinate and body coordinate through human place under different statuses of predefined acceleration and angular velocity information in the human body. Besides, Huiyu Zhou (Zhou et al., 2008) adopted body skeleton model linked by the human joint and got the space position between arm elbow and wrist joints through the transformation between the coordinate system. Even though the method can locate the joint position, it cannot get the corresponding flip Angle of arm turning.