Face Image Encryption Using Fuzzy K2DPCA and Chaotic MapReduce

Abstract

As technology continues to advance, safeguarding personal privacy and information security has become increasingly critical. Facial image encryption algorithms involve encrypting and decrypting facial images to prevent unauthorized access and malicious use. Fuzzy computation is a practical solution for many decision problems in facial recognition encryption algorithms. To this end, this study proposes the use of fuzzy two-dimensional kernel principal component analysis for facial recognition and chaotic MapReduce for facial image encryption. The study introduces fuzzy membership functions to handle uncertainty in two-dimensional kernel principal component analysis. Experimental results indicated that the accuracy of the fusion fuzzy calculation and two-dimensional kernel principal component analysis method exceeded 75%, which was 13-37% higher than the comparison method. Furthermore, the proposed method combining chaotic systems and MapReduce has a uniform histogram distribution and runs 50% faster than the comparison method. Consequently, it is evident that the method proposed by the research institute is both feasible and efficient for safe facial image analysis.

Keywords

• encryption algorithm
• facial image recognition
• fuzzy computing
• two-dimensional kernel principal component analysis