Abstract
In this paper, we propose a gait analysis method to extract the dynamic and static information from the input video for walking path determination and human identification. Based on the periodicity of swing distances, we may estimate the gait period of each walking video sequence. For each gait cycle, we depict the dynamic information by analyzing the distribution of motion vectors, and then describe the static information by using Fourier descriptors. The extracted dynamic and static information is transformed into lower dimensional embedding space for human identity recognition. To solve the difference of walking velocity between the test and training human objects, a hybrid human ID recognition algorithm is developed to choose the effective feature. Given a test feature vector, the nearest neighbor classifier is applied for walking paths determination and human identification. The proposed algorithm is evaluated on the CASIA gait database, and the experimental results demonstrate a highly acceptable recognition rate, for example, 98% for normal walking dataset.
Paper version not known (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call