Modeling and Reliability Enhancement of Access Subject Identification in PACS Based on Hybrid PCA Algorithms

Abstract

The analysis of current challenges accompanying the effective operation modern Physical Access Control Systems (PACS) reveals their key problems: a significant increase in the volume of identification features, which overloads traditional methods, and high sensitivity to non-standard reading conditions, leading to errors. The PCA method effectively addresses the problem of data dimensionality reduction and accelerates search. However, this method remains vulnerable to non-ideal conditions, limiting its application and reliability in dynamic PACS environments. To systematize and gain a detailed understanding of the identification process, a structured approach with modeling using the IDEF0 methodology was applied, which highlighted the critical importance of comparing the transformed image with reference templates. The integration of PCA method with Convolutional Neural Networks, Machine Learning, and Deep Learning algorithms was analyzed based on the detection of hidden nonlinear patterns and robust features. Industry standards such as FAR (False Acceptance Rate), FRR (False Rejection Rate) and EER (Equal Error Rate) were used to evaluate the effectiveness of hybrid methods. The research results confirm the relevance of hybrid approaches for ensuring maximum security and operational efficiency of modern PACS