Year: 2024
Type: article
Abstract: In modern society, the importance of facial encryption technology has become increasingly evident. However, existing encryption techniques generally suffer from a significant issue: excessive encryption levels create too much noise, which not only attracts the attention of hackers but may also impair the judgment of facial movements. To address this issue, this paper proposes a detail encryption system which contains two parts: (1) facial detection based on the Viola–Jones object detection framework, and (2) AES-supported detail encryption. The facial detection part aims at localizing the facial features which are appropriately encrypted. Experimental comparisons show that the face images processed by the detail encryption technique obscure distinct facial features. This finding indicates that while the detail encryption technique significantly reduces facial recognizability, it does not overly impair the perception of facial movements.
Author: Xin Liu
Institution: University of Science and Technology Beijing
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4405592762