A hand gesture recognition method, system, device and storage medium

Abstract

The present application relates to a hand gesture recognition method, system, device and storage medium. The method includes: capturing the RGB image of the hand from the RGB camera, capturing the depth image of the hand from the active depth camera, and according to the stereo image Obtain a hand pose data set with the depth image; process the hand pose data set to obtain a 3D joint position, and use the 3D joint position as a data set to mark the training of the software model; The feature extractor extracts the RGB image to obtain the feature map of the hand posture; process the feature map according to the attention mechanism to obtain the global feature map of the hand posture, and use the global feature map to obtain the recognition of the hand posture result. While providing recognition accuracy, the recognition method verifies the effectiveness of the coordinate attention mechanism module and the multispectral attention mechanism module in the hand pose estimation network feature extractor.

Description

technical field [0001] The present application relates to the field of artificial intelligence, in particular to a hand gesture recognition method, system, device and storage medium. Background technique [0002] With the continuous development of the information age, various electronic devices equipped with artificial intelligence information systems have been integrated into our lives, and the demand for human-computer interaction has become increasingly prominent. As the basic work of hand interaction, human hand position recognition positioning and pose estimation also have great research value. There are many approaches for hand pose estimation using deep learning. There are methods based on image plus depth information, methods based directly on RGB images, and methods based on binocular or even multi-eye vision. Among them, the hand pose estimation method based on a single RGB image has attracted more attention because of its simple implementation, low hardware requ...

AI Technical Summary

Problems solved by technology

However, the above recognition process has the problem of low recognition accuracy

Images