Attitude detection method and device, electronic equipment and computer readable storage medium

Abstract

The embodiment of the invention provides a posture detection method and device, electronic equipment and a computer readable storage medium, and relates to the field of computer processing. The method comprises the following steps: acquiring an image sequence corresponding to a target object and motion data of a target part of the target object, and determining first three-dimensional coordinate information of each key point of the target part in any image in the image sequence; and determining a first unit vector representing the direction of the target part according to the first three-dimensional coordinate information of each key point. And determining a second unit vector representing the direction of the target part based on the motion data of the target part. And determining the attitude of the target part according to the similarity between the first unit vector and the second unit vector and the first three-dimensional coordinate information of each key point. According to the method, the first three-dimensional coordinate information of the target part is combined with the motion data of the target part, so that the posture of the target part can be accurately determined, and the user perception is improved.

Description

technical field [0001] The present application relates to the field of computer processing, and in particular, the present application relates to a gesture detection method, apparatus, electronic device, and computer-readable storage medium. Background technique [0002] With the development of deep learning, human pose estimation has higher and higher accuracy, and more applications have emerged. For example, some terminal devices support somatosensory games, AI (Artificial Intelligence, artificial intelligence) fitness, etc., and some live broadcast software provide 3D (3-dimension, three-dimensional) virtual character drivers. [0003] In the related art, the deep learning algorithm based on monocular vision can adapt to the vast majority of scenes, but when some parts of the organism are occluded or the picture is blurred due to the rapid movement of the limbs, the human posture obtained based on the deep learning algorithm can be obtained. The accuracy is poor, and eve...

AI Technical Summary

Problems solved by technology

[0003] In related technologies, the deep learning algorithm based on monocular vision can adapt to most scenes, but when some parts of the organism are occluded or the picture is blurred due to rapid movement of the limbs, the human body posture obtained based on the deep learning algorithm The accuracy is poor, and there may even be obvious jitter

Although the use of inertial sensors can solve the problem of poor posture accuracy due to occlusion of certain parts of the organism, the measurement values ​​of the inertial sensors will have errors, and the measurement values ​​of the inertial sensors are accumulated to calculate the displacement of the organism, lead to a noticeable shift

It can be seen that none of the above methods can detect the pose of a more accurate organism.

Images