Gesture feature acquisition and recognition method and system for manned cockpit

By using adaptive lighting control and deep learning technology in an aviation environment for gesture acquisition and recognition, the problem of low gesture recognition accuracy in complex environments has been solved, and an efficient and reliable gesture recognition system has been achieved.

CN122336837APending Publication Date: 2026-07-03AEROSPACE LIFE SUPPORT IND LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610317520.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-16
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing aviation gesture recognition systems have low accuracy in complex aviation environments and cannot meet aviation-grade reliability standards.

Method used

It employs an adaptive lighting environment control system for the fill light component and lens aperture, combined with deep learning and image processing technologies, to perform gesture acquisition, enhancement, noise reduction, segmentation, and feature fusion. It utilizes depth images and color space to extract gesture regions, and combines 3D convolutional networks and graph convolutional networks for recognition and classification.

Benefits of technology

High-accuracy gesture recognition was achieved in complex aviation environments, improving operational efficiency, reducing the false recognition rate, and enhancing the robustness and adaptability of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122336837A_ABST
    Figure CN122336837A_ABST
Patent Text Reader

Abstract

This application discloses a method and system for gesture feature acquisition and recognition in manned aircraft cockpits, including the following steps: gesture acquisition and processing: adaptively controlling the supplementary lighting component and lens aperture according to the lighting environment to acquire a clear original image of the pilot's hand; enhancing, denoising, and performing analog-to-digital conversion on the original image to generate a digital image signal; data preprocessing: preprocessing the digital image signal to remove noise; gesture segmentation: based on color and depth information, first extracting the gesture region from the depth image, then performing gesture color segmentation in the color space, and then extracting the gesture contour; feature fusion: extracting gesture features including fingertip position, number of fingertips, fingertip angle, and palm center from the gesture contour; recognition result classification: inputting the gesture features into a deep learning-based classification model to obtain the gesture category. This application ensures recognition accuracy in aviation environments.
Need to check novelty before this filing date? Find Prior Art