Human coxitis gait feature detection method based on gait analysis and attitude estimation
By combining gait analysis and pose estimation methods, using video sequences to extract gait characteristics and using LightGBM network for detection, the cost and non-automated problems of automated detection and analysis of gait diseases in the prior art are solved, and efficient and accurate automated analysis and scoring of gaits in patients with hip arthritis are achieved.
CN119723666APending Publication Date: 2025-03-28NANJING UNIV OF POSTS & TELECOMM
13 Cites 0 Cited by
Patent Information
- Application Number
- CN202411856174.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-17
- Publication Date
- 2025-03-28
AI Technical Summary
Technical Problem
The prior art has high cost and non-automated problems in automated detection and analysis of human gait diseases, especially in applications in the medical field, which is difficult to achieve automated analysis.
Method used
Using a method based on gait analysis and pose estimation, gait continuity features are extracted through video sequences, combined with RTMPose key point detection algorithm and LightGBM network, automated detection and scoring of gait characteristics of patients with hip arthritis are achieved.
Benefits of technology
It effectively reduces labor costs, realizes automated analysis and scoring of gaits for patients with hip arthritis, and improves the accuracy and efficiency of detection.
✦ Generated by Eureka AI based on patent content.
Abstract
The invention discloses a human coxitis gait feature detection method based on gait analysis and attitude estimation. The method comprises the following steps: processing a gait video into a gait image sequence; key point extraction is carried out on a walking object in the gait image sequence, and a gait key point sequence is formed; gait features of a normal subject and a patient with coxitis can be obviously distinguished through design and calculation; the linear correlation between the gait features is verified by using a Pearson's correlation coefficient, and the gait features with relatively high correlation are selected; and sending the selected gait features into a LightGBM network for training, adjusting and optimizing parameters, and finally distinguishing normal subjects and patients with coxitalitis according to the gait features, namely realizing coxitalitis detection in the gait video. According to the method, a method of combining gait analysis and posture estimation is effectively utilized, the gait motion characteristics of the patient suffering from coxitalitis are fully excavated, and the gait score of the patient suffering from coxitalitis is effectively calculated.
Need to check novelty before this filing date? Find Prior Art