Research On Gait Recognition Method Based On Deep Learning

With the development of science and technology,identification technologies based on various biological features such as iris,face and voiceprint have become more and more popular.One of the most attractive biometric technologies at this stage is gait recognition.However,gait recognition technology still has many problems,such as complex scenes,insufficient cross-view feature extraction,occlusion of pedestrians or coats.Aiming at the problem of occlusion of pedestrians and insufficient gait extraction under cross-view,this paper proposes a gait recognition method based on gait contour map and improved convolutional neural network(CNN),through deep learning.Convolutional neural networks,which are very advantageous in image processing,perform similarity learning.The method was tested on the published gait data set CASIA-B,and compared with the existing methods,the results show that the accuracy of the method is higher than other methods.Based on the proposed gait contour map based on gait contour map and improved convolutional neural network(CNN),this paper proposes a gait recognition method based on gait image sequence and CNN-LSTM,which is better.The problem of processing the gait time series,while masking the input gait sequence diagram,makes the improved method show better robustness in the case of pedestrian clothing and carrying items.Finally,experiments were carried out on the open gait dataset CASIA-B.Experiments show that the proposed method has improved recognition accuracy compared to other existing methods and previously proposed methods.