A human pose prediction method based on Gaussian process regression and progressive filtering

By using a method based on Gaussian process regression and progressive filtering, a position state transition model for human joints is established, which solves the uncertainty problem in human posture prediction in traditional methods and achieves high-precision posture prediction in complex environments.

CN115050095BActive Publication Date: 2026-06-12ZHEJIANG UNIV OF TECH
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ZHEJIANG UNIV OF TECH
Filing Date
2022-06-06
Publication Date
2026-06-12

AI Technical Summary

Technical Problem

Traditional methods struggle to accurately predict human pose in complex scenes and under motion conditions. In particular, the sensors are susceptible to uncertainties caused by ambient lighting and occlusion, and noise suppression is ineffective. Existing technologies require extensive human image annotation and have limited accuracy.

Method used

By employing a method based on Gaussian process regression and progressive filtering, a position state transition model of human joints is established. Attitude prediction is performed using sensor measurements, and the attitude is corrected within the confidence interval, thereby improving the accuracy and robustness of attitude prediction.

Benefits of technology

It effectively improves the accuracy and robustness of human posture prediction, reduces the impact of sensor errors on prediction results, and enhances the accuracy of target tracking.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115050095B_ABST
    Figure CN115050095B_ABST
Patent Text Reader

Abstract

The present application relates to the field of human posture prediction, especially to a human posture prediction method based on Gaussian process regression and progressive filtering, which uses the Gaussian process regression method to establish a state transition model, uses the predicted value to test the reliability of the measured information, to improve the accuracy of the prior information; then estimates the position state based on the progressive filtering method, to achieve faster and more accurate performance; finally, the position state of each joint of the human body is predicted. The beneficial effects of the present application are: due to the fact that the sensor is easily affected by environmental factors and its own influence when obtaining human posture information, which may lead to the loss of measured information or obvious errors, thus making the human posture prediction result deviate obviously, compared with the existing human posture prediction, this method is based on Gaussian process regression and progressive filtering, which effectively improves the accuracy of state prediction and improves the robustness to adverse factors such as sensor errors.
Need to check novelty before this filing date? Find Prior Art