Method for evaluating gait recognition performance based on motion capturing frame

Abstract

The invention provides a method for evaluating gait recognition performance based on a motion capturing frame. The method includes the main steps that a database is established; gait features are extracted; class separation degrees are evaluated, and the frame is evaluated, wherein gait data is collected to establish the database GaitMo, through Fisher linear discriminant analysis improved based on the maximum margin criterion (MMC) and in combination with principle component analysis (PCA) and linear discriminant analysis (LDA), the gait features are learned, and finally according to a specific application scene, appropriate indexes are selected from separation coefficients of four classes of feature space and four classifier performance indexes based on sequence to carry out evaluation so as to evaluation a system. The limit that new methods emerge endlessly and have difficulty in comparison and analysis is broken through; by providing the evaluation frame, set evaluation indexes are evaluated on a reference database, in this way, research, evaluation and analysis become effective, simple and convenient for gait recognition, contributions are made for repeatability research, and research and development of gait recognition are promoted.

Description

technical field [0001] The invention relates to the field of biological recognition, in particular to a method for evaluating gait recognition performance based on a motion capture framework. Background technique [0002] Gait recognition is often used in the fields of intelligent video monitoring, walking function evaluation, performance evaluation of rehabilitation equipment, etc. Among them, in intelligent video monitoring, it has more advantages than face recognition, and it is recognized by people's walking posture. The biometric identification technology of identification has the advantage that it is difficult to be artificially hidden or disguised. In recent years, the urgent needs of anti-terrorism, information security, rescue, medical treatment, and people's livelihood have made gait recognition technology one of the active research directions in the field of biometric identification, and new methods have emerged in an endless stream. In order to promote this highl...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that existing methods need more effective comparison to promote research, the purpose of the present invention is to provide a method for evaluating gait recognition performance based on the motion capture framework, and to learn the extraction of gait features by two methods: MMC and PCA+LDA , evaluate the class separability and then evaluate the system, contribute to the repeatability research, and promote the research and development of gait recognition

Images