Gesture recognition method and device, electronic equipment and readable storage medium

Abstract

The embodiment of the invention provides a gesture recognition method and device, electronic equipment and a readable storage medium, and relates to the field of image processing. The method comprises the steps of obtaining gesture posture estimation data of an image to be recognized, wherein the gesture posture estimation data comprise position information of a first joint point of a hand; fitting a preset first hand model based on the gesture estimation data to obtain a second hand model; according to node information of each second joint point in a second hand model, adjusting the gesture attitude estimation data to obtain target attitude data conforming to the structure of the second hand model; and performing gesture recognition according to the target posture data to obtain a recognition result. Therefore, the position information of each joint point conforming to the actual hand model structure can be obtained, the position information of the joint points used during gesture recognition is more accurate, drifting of points can be reduced, the forming rate of the hand shape is increased, and therefore the accuracy of the recognition result is improved.

Description

Technical field [0001] The present application relates to the field of image processing, and in particular, there is a gesture identification method, a device, an electronic device, and a readable storage medium. Background technique [0002] At present, when gesture recognition is performed, it is generally first to obtain a critical point position of the hand in the image in the image, and then the gesture recognition model is used to identify the identification of the gesture based on the key position. However, when the gesture recognition model is used to obtain a critical point position, it may be not accurate due to the impact of various factors, which is not accurate enough from the image, and the drift is too large, which in turn leads to the position of the key point. The gesture recognition result is low accuracy. Inventive content [0003] The present application embodiment provides a gesture identification method, device, electronic device, and readable storage mediu...

AI Technical Summary

Problems solved by technology

However, when using the gesture recognition model to obtain the key point position, due to the influence of various factors, the key point position obtained from the image may not be accurate enough, and the drift relative to the real position is too large, which in turn leads to the result based on the key point position. The accuracy of gesture recognition results is low

Images