Key point detection method of human body posture image

Abstract

The invention provides a key point detection method of a human body posture image. By using an estimation network model provided by the invention, the characteristics are divided into different channels by designing the segmentation channel blocks, so that the image characteristics become richer, and the accuracy of human body posture detection can be effectively improved. Besides, an attitude correction machine is designed, and by improving a channel attention mechanism and a space attention mechanism, a context attention mechanism is introduced to improve the modification effect. The accuracy of human body posture estimation can be effectively improved.

Description

Technical field [0001] The invention relates to the field of image processing, and in particular to a method for detecting key points of human posture images. Background technique [0002] Cascaded Pyramid Network (CPN) is a coarse-to-fine network model that uses context information of a single person to complete human pose estimation. Regional Multi-Person Pose Estimation (RMPE) consists of three parts: Symmetric Space Transformation Network (SSTN), Parametric Pose Non-Maximum Suppression (NMS) and Pose Guided Proposal Generator (PGPG). It processes inaccurate human positioning frames and Redundant detection effectively improves the accuracy of human posture estimation. These human body estimation algorithms effectively improve the accuracy of model recognition, but there are problems such as the model is too large and the prediction speed is slow, which is not conducive to practical use. Contents of the invention [0003] The present invention provides a key point dete...

AI Technical Summary

Problems solved by technology

These human body estimation algorithms have effectively improved the accuracy of model recognition, but there are problems with too large models and slow prediction speeds, which are not conducive to practical use.

Images