Hand key point detection method and device, and computer equipment

Abstract

The invention relates to a hand key point detection method, a hand key point detection device and computer equipment. The key point detection method comprises the following steps of: acquiring an image set; generating masks of hand key points according to real position labels of the hand key points in the image set, wherein the hand key points are classified into fingertip points and a palm centerpoint; according to a feature map of the image set, analyzing key point identification information of the hand key points; according to the masks and the key point identification information, acquiring predicted position label values of the hand key points; according to the predicted position label value and the real position labels, carrying out loss function regression on the hand key points toobtain a hand key point detection model; and conducting hand key point detection by using the hand key point detection model. Through adopting the hand key point detection method and the hand key point detection device, the problem that a method for carrying out posture estimation depending on orderly-arranged key points is high in cost in the prior art is solved, the scene adaptability of hand key point detection is improved, and the cost of hand key point detection is reduced.

Description

technical field [0001] The present application relates to the field of computer vision, in particular to a method, device and computer device for hand key point detection. Background technique [0002] With the development of human-computer interaction technology, human-computer interaction methods such as keyboards, mice, and touch screens have been difficult to meet user requirements in many emerging fields. In remote control scenarios such as augmented reality (Augmented Reality, referred to as AR), users It is more inclined to use wireless and non-contact methods to realize human-computer interaction. On the other hand, the interactive technology based on computer vision technology can allow users to get rid of complicated interactive devices, and use specific body movements to send instructions to the machine, which is convenient and fast. way to meet the requirements of human-computer interaction. [0003] In related technologies, the attitude estimation of key point...

AI Technical Summary

Problems solved by technology

[0005] Embodiments of the present application provide a method, device, computer device, and computer-readable storage medium for hand key point detection, so as to at least solve the problem of relatively high cost in related art methods that rely on orderly arranged key points for pose estimation. The problem

Images