Facial Data Classification Through Enhanced Local Binary Patterns (LBP) and Dynamic Range Local Binary Patterns (DRLBP) Algorithms

Abstract

A significant amount of reliance is placed on facial data classification in contemporary computer vision and pattern recognition. This research presents a novel method that makes use of the algorithms of Dynamic Range Local Binary Patterns (DRLBP) and Enhanced Local Binary Patterns for the purpose of face data classification that is both effective and precise (LBP). The classic LBP methodology is expanded upon by the Enhanced LBP method, which incorporates adaptive thresholding techniques and spatial histogram characteristics. This makes it possible to conduct a more thorough investigation of the texture and to be resilient in a variety of lighting conditions. By continuously modifying the local binary pattern range, the DRLBP algorithm improves upon this in order to better accommodate nuanced facial features and expressions. This is done in order to better accommodate facial expressions. In terms of accuracy, speed, and adaptability, our proposed system beats state-of-the-art alternatives, as demonstrated by extensive trials conducted on commonly used facial datasets. According to the findings of our investigation, it would appear that human-computer interaction (HCI), digital forensics, and security systems could all stand to gain a great deal from a solution that combines Enhanced LBP and DRLBP algorithms for the classification of face data.