Automatic wrinkle detection and wrinkle evaluation

Wrinkle is a kind of important facial feature widely concerned in facial-image-based applications, such as age estimation, face inpainting, skin quality assessment and so on. Obviously, the extraction of wrinkle features is an essential task in these application scenarios. However, existing algorith...

Full description

Saved in:
Bibliographic Details
Main Author: Zhang, Yuxian
Other Authors: Alex Chichung Kot
Format: Thesis-Master by Coursework
Language:English
Published: Nanyang Technological University 2021
Subjects:
Online Access:https://hdl.handle.net/10356/153213
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Wrinkle is a kind of important facial feature widely concerned in facial-image-based applications, such as age estimation, face inpainting, skin quality assessment and so on. Obviously, the extraction of wrinkle features is an essential task in these application scenarios. However, existing algorithms of wrinkle detection focus more on coarse horizontal wrinkles in the forehead area, which may not satisfy most of the application requirements. In this case, this dissertation creates an end-to-end process of wrinkle detection and evaluation, realizing the extraction of more types of wrinkles. Based on the concept of wrinkle index, evaluation of wrinkle severity is given in the form of a specific score, giving support to skin quality assessment. Results show that most of the visible wrinkle patterns can be successfully extracted, and the evaluation score also matches with human perception of wrinkle severity. This work would be helpful in the field of computer-based automatic detection.