Fractal encoding uses and variations in contemporary biometrics applications
Carmen Bisogni, Aniello Castiglione, Fabio Narducci, Chiara Pero · 2023
Fractal encoding, as part of Computer Vision processes, has gained major success as the base of a lossy compression algorithm. Yet, the principle of self-similarity, which under-pins fractal encoding, has sparked attention in the Biometrics research field throughout the years. In contrast to the era of Deep Learning techniques, fractal encoding and its integration with Machine Learning produced intriguing and competitive outcomes in real-world use applications. The goal of this article is to gather biometrics applications that have used the fractal encoding approach to tackle authentication, recognition, or classification problems in these sectors in recent years. We investigate how the authors adopted fractal encoding to be a main feature in a biometric system in a reasonable encoding computational time. The approaches for classifying and regressing fractal encoding to get the desired output are described, and the more promising results are emphasized. After a review of the methodologies previously researched in this field, we outline a direction for future research on fractal encoding applied to biometrics. The findings achieved by the approaches discussed in this study show that there is still a large and potential scope for research in this field.