Gait Recognition Based On The Plantar Pressure Image And Sparse Representation

With the rapid development of computer technology, the information age brought us convenience, but there are also a lot of security vulnerabilities. In view of the the traditional identification technology can not meet the demand of the people of the safety. This paper describes the gait recognition technology.The pressure information that the plantar pressure image acquisition equipment got is a three-dimensional gait image time series information. At first this research transformed three-dimensional information into a two-dimensional feature information. Subsequently, measured the two-dimensional characteristics by linear sparse representation on the training set. Each category recovery two-dimensional information by the sparse coefficient. Measured the difference of actual two-dimensional characteristics and recovery two-dimensional characteristics, we think that the category is the smallest difference class. Of course, the sparse representation coefficient should meet the requirements of the sparse concentration index.This paper proposes a new feature of plantar pressure image, mean plantar pressure image features. The relationship between the existing feature and the new feature is complementary. The fusion of the new feature and maximum plantar pressure image feature obtains good recognition as the characteristics as the input. This paper is the first time to use sparse representation on gait recognition which based on plantar pressure image. And it proved this method is very suitable for plantar pressure image-based gait recognition using gait recognition on the plantar pressure image.