A human body posture estimation method and mobile terminal

Abstract

The invention discloses a human body pose estimation method, which is suitable for marking human body key points in a video to be processed to extract human body pose information, and is suitable for execution in a mobile terminal, comprising the steps of: acquiring the video to be processed, and Determine whether the key point position of the previous image frame of the current image frame exists; if there is no key point position of the previous image frame, then detect the position of the human body in the current image frame, and generate a target frame containing the position of the human body; if there is a key point position of the previous image frame The key point position of an image frame, then generate the target frame corresponding to the current image frame according to the key point position of the human body marked in the previous image frame; perform scale transformation on the generated target frame to generate a human body image of predetermined size; The human body image is input to a predetermined pose estimation model to output an image marked with key point positions of the human body; and the human body pose is estimated according to the marked key point positions. The invention also discloses the corresponding mobile terminal.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a human body pose estimation method and a mobile terminal. Background technique [0002] Human body pose estimation refers to the process of labeling the key points of the human body in the image or video, and then extracting the basic pose information of the human body, such as figure 1 As shown in , the dots represent the marked key points of the human body, and the basic posture information of the human body can be obtained according to the marked key points of the human body. Commonly used key points of the human body (that is, human body joint points or skeletal points) include the head, neck, shoulders, elbows, wrists, hips, knees, and ankles. Human pose estimation is to identify the positions of these key points on the human body in the image video. [0003] Existing human pose estimation algorithms can be divided into traditional vision methods and deep learnin...

AI Technical Summary

Problems solved by technology

[0004] However, in actual scenes, due to the different proportions of the human body in different images or videos, different clothing, occlusion relations and complex shooting angles, it is often difficult to estimate the human body pose quickly and effectively.

In addition, in addition to the diversity of target appearance, the efficiency of algorithm operation is also an issue that needs to be considered

The current pose estimation algorithm cannot meet the requirements of mobile applications in terms of computational efficiency.

Images