Method for quantifying metabolic chamber animal behavior phenotypes
By using an RGB-D side-view acquisition system and deep learning algorithms, the problems of anti-interference and accuracy in the complex environment of the metabolic chamber were solved, and high-precision behavior quantification and automated analysis were achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-23
- Publication Date
- 2026-07-17
AI Technical Summary
In the complex environment of the metabolic chamber, existing technologies have weak anti-interference capabilities, limited spatial resolution, and insufficient accuracy in motion state measurement, leading to deviations in behavior quantification results.
An RGB-D side-view acquisition system is used, combined with deep learning algorithms, to segment animal target images through a multi-level filtering algorithm with depth threshold and region constraints. High-precision behavior recognition is achieved by automatically analyzing the key points of three-dimensional skeletons and posture features.
It achieves millimeter-level precision in physical space motion quantization, significantly improves anti-interference capabilities, accurately identifies fine behaviors in complex environments, and realizes full-process automation and high-precision analysis.
Smart Images
Figure CN122415536A_ABST