查看文章

The fingerprint-based authentication systems are being extensively deployed as security tool for providing access to the critical Cyber Physical Systems (CPS). However, this rapid relocation to fingerprint-based recognition poses many security concerns that are primarily due to presentation or spoofing attacks on sensor module. To alleviate these attacks, presentation attack detection (PAD) mechanisms are developed that mainly rely on micro-textural properties of the fingerprint images. In this paper, we expound a novel intelligent fingerprint PAD (IFPAD) approach for securing typical CPS that exploits two micro-textural features from an image. We propose a new Local Adaptive Binary Patterns (LABP) descriptor that extracts more refined local information by using dynamically adapted threshold. Moreover, to augment more discriminative power, the Uniform Local Binary patterns (ULBP) descriptor is coalesced with …