Face key point determination method and device and storage medium

Abstract

The embodiment of the invention discloses a face key point determination method and device and a storage medium. By updating the position coordinate data of each key point region in the first positioncoordinate set of the current frame, the jitter of the face key points of each key point region can be reduced, and the jitter reduction accuracy is realized. The relative offset capable of representing the overall movement amount of the target key point area between the current frame and the previous frame is obtained through calculation, so that the position coordinate data of the target humanface key point in the updated first position coordinate set can be accurately determined by utilizing the relative offset, and the jitter of the human face key point is further reduced.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method, device and storage medium for determining human face key points. Background technique [0002] The function of face key points is to accurately locate and segment the position of various parts of the face, such as the precise outline of eyes, eyebrows, mouth, the outline of the face, etc., and there are many fields that face key points can be applied to, such as human Face deformation (thin face, big eyes, etc.), virtual makeup and animation movies and other fields. [0003] In the dynamic virtual technology based on face video, it is usually necessary to detect the key points of the face. However, the accuracy and stability of the existing key point detection methods of the face are low, and the detected key points of the face are easy to produce Vibration in visual effects. Contents of the invention [0004] The main purpose of the present inventi...

AI Technical Summary

Problems solved by technology

[0003] In the dynamic virtual technology based on face video, it is usually necessary to detect the key points of the face. However, the accuracy and stability of the existing key point detection methods of the face are low, and the detected key points of the face are easy to produce Vibration in visual effects

Images