Walking feature dimension reduction method for human-shaped intelligent agent three-dimensional behavior space

Abstract

The invention discloses a walking feature dimension reduction method for a human-shaped intelligent agent three-dimensional behavior space. The method comprises the following steps: two-dimensional consideration analysis, establishment of a geodetic coordinate system, step detection analysis, trunk motion state analysis during walking and leg motion state analysis during walking. In complex multi-dimensional spatio-temporal data processing, the position of a human-shaped intelligent agent is analyzed, considered and calculated by adopting a method of reducing dimensions into two dimensions; burden on computer operations is alleviated, and a northeast sky coordinate system is established; the instant position of a user is obtained through the initial position of a starting point and the advancing direction and distance of each step, and subsequent systematic errors are reduced; meanwhile, an accelerometer is used for step detection, whether each step is completed or not in the human-shaped intelligent detection hiking process is detected, whether steps are generated or not is determined, whether the steps move forwards or backwards or slide leftwards or rightwards is determined by analyzing acceleration signals, the situation of step loss or multiple steps is avoided, and the data accuracy is guaranteed.

Description

technical field [0001] The invention relates to the technical field of walking feature analysis, in particular to a method for dimensionality reduction of walking features in a three-dimensional behavior space of a humanoid agent. Background technique [0002] Humanoid agent position estimation is generally divided into steps: step detection, step size estimation and heading angle estimation. Once the step length and heading angle are determined, the final position after the current step is completed can be released; [0003] In complex multi-dimensional spatio-temporal data processing, the computational burden of the computer is heavy, the walking characteristics analyzed are not comprehensive enough, and there are system errors or detection errors, which lead to step loss and multiple steps during the traveling process, which affects the accuracy of the data. Therefore, this paper The invention proposes a walking feature dimensionality reduction method in the three-dimens...

AI Technical Summary

Problems solved by technology

[0003] In complex multi-dimensional spatio-temporal data processing, the computational burden of the computer is heavy, the walking characteristics analyzed are not comprehensive enough, and there are system errors or detection errors, which lead to step loss and multiple steps during the traveling process, which affects the accuracy of the data. Therefore, this paper Invention proposes a walking feature dimensionality reduction method in the 3D behavior space of humanoid agents to solve the problems existing in the prior art

Images