Human stillness / posture determination method based on 1t2r feature abstraction

By employing the 1T2R feature abstraction method and utilizing the difference features of dual receiving channels and time stability analysis, the problem of determining the static state and attitude of the human body under low-resolution conditions in millimeter-wave radar is solved. This achieves low-power, low-cost, and high-accuracy human body perception, which is suitable for smart home and other scenarios.

CN122172183APending Publication Date: 2026-06-09谭鑫
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
谭鑫
Filing Date
2026-02-26
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing millimeter-wave radars have difficulty accurately determining the static state and basic posture of the human body under low-resolution conditions, and they also suffer from high power consumption and computational complexity, which is particularly difficult to achieve in low-power, long-running IoT devices.

Method used

The 1T2R feature abstraction method is adopted, which utilizes the difference features of dual receiving channels and time stability analysis. By using one transmitting antenna and two spatially separated receiving antennas, the difference features of the receiving channels are extracted and the stability of changes is analyzed within a preset time window to generate a static existence score. The basic attitude is determined by combining the equivalent longitudinal expansion index, thus avoiding the generation of radar point cloud data and complex recognition algorithms.

Benefits of technology

It achieves accurate determination of the static presence and basic posture of the human body under low power consumption conditions, reduces device hardware costs and computational complexity, avoids privacy leakage risks, is suitable for low-cost embedded platforms, and ensures real-time performance and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122172183A_ABST
    Figure CN122172183A_ABST
Patent Text Reader

Abstract

This invention discloses a method for determining human stillness / posture based on 1T2R feature abstraction, relating to the field of millimeter-wave radar technology. Specifically, it is a method for determining human stillness / posture based on 1T2R feature abstraction. The method includes: acquiring echoes using 1T2R millimeter-wave radar; extracting differential features based on the geometric differences between dual receiving channels; calculating a stillness score based on the stability of feature changes within a time window to distinguish between unoccupied, still, and slightly moving states; and determining basic postures such as lying down, sitting, and standing through feature spatial distribution or equivalent longitudinal expansion indices. The system includes radar, feature extraction, stillness scoring, and posture determination modules, without point cloud output. This method does not rely on high-resolution point clouds, complex algorithms, or deep learning, does not identify identities, or output final behavioral conclusions, achieving low-complexity human stillness and basic posture determination.
Need to check novelty before this filing date? Find Prior Art