Gait abnormality classification method based on deep convolution neural network

Abstract

The invention belongs to the technical field of biometric recognition, and particularly relates to a gait abnormality classification method based on a deep convolution neural network. The method includes the following steps that a IMU worn on human body is used for collecting signals of normal walking and simulating typical abnormal gait walking, and triaxial acceleration information under different gaits is obtained; according to the typical walking frequency of a target, original data is subjected to windowing cutting pretreatment, and each data queue is correspondingly labeled according tothe gait type, wherein the CNN deep convolution neural network includes a first convolution layer, a second convolution layer, a first pooling layer, a second pooling layer, a full connection layer and a soft max output layer; finally, data labels are divided into a training set and a test set, the training set is sent to the CNN for training, and the test set is used for evaluating the model classification effect after training. The method omits complicated gait cycle division and feature extraction engineering, improves the classification accuracy of various abnormal gaits, reduces the workload of data preprocessing, and improves the classification accuracy.

Description

technical field [0001] The invention belongs to the technical field of biometric feature identification, and in particular relates to an abnormal gait identification method. Background technique [0002] Gait refers to the posture shown when people walk, and is one of the important biological characteristics of the human body. Abnormal gait is mostly related to the lesion site. As an important feature reflecting the health status and behavior ability of the human body, it has attracted much attention in clinical research such as medical diagnosis and disease prevention. [0003] The current mainstream gait recognition methods are mainly divided into computer vision solutions based on video and image processing and sensor solutions based on walkways and wearable sensors such as IMU. Both schemes involve a large amount of professional data preprocessing and complicated feature engineering after obtaining the original data in order to extract the relevant features in the gait ...

AI Technical Summary

Problems solved by technology

Both schemes involve a large amount of professional data preprocessing and complicated feature engineering after obtaining the original data in order to extract the relevant features in the gait cycle. Although the accuracy is high, the real-time performance is poor and the computational complexity is high.

The mainstream gait recognition system only provides various numerical indicators, and the identification and classification tasks of abnormal gait are mainly completed by human experts, which requires a lot of professional knowledge in related fields

Images