China
Africa
Explore chapters and articles related to this topic
Introduction to Biometry
Published in Ling Guan, Yifeng He, Sun-Yuan Kung, Multimedia Image and Video Processing, 2012
Carmelo Velardo, Jean-Luc Dugelay, Lionel Daniel, Antitza Dantcheva, Nesli Erdogmus, Neslihan Kose, Rui Min, Xuran Zhao
Soft biometrics are physical, behavioral, or adhered human characteristics classifiable in predefined human-compliant categories. These categories are, unlike in the classical biometric case, established and time proven by humans with the aim of differentiating individuals. In other words, the soft biometric trait instances are created in a natural way, used by humans to distinguish their peers.
Optimal Score Level Fusion for Multi-Modal Biometric System with Optimised Deep Ensemble Technique
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2023
M. R. Bharath, K. A. Radhakrishna Rao
Biometric devices are one of the most extensively utilised forms of human authentication in the world. The introduction of devices capable of acquiring these biometric traits, such as profile, iris, fingerprint, palmprint, soft biometrics, palm vein and finger vein, has benefited the advancement of biometric authentication research in both unimodal and multi-modal ways (Adersah and Kalkur 2017; Asderah 2019; Walia et al. 2020; Vijay and Indumathi 2021; Banchini et al. 2022). Multi-modal systems are typically thought to be more safe than unimodal systems because they are more stable and reliable (Tiong et al., 2020). As a result, vein images from various hand locations, such as the palm and back of the palm (dorsal), as well as the elbow, are the most desirable biometric features, as these vascular veins beneath the skin are extremely difficult to spoof (Li et al., 2020). Additionally, inexpensive sensors that can gather vein patterns from different hand locations have grown in popularity for the application in high-security detection methods (Kumar et al. 2019). The usage of such devices is much more easily accepted by customers, thanks to the image capture technique’s extreme simplicity, safety and ease of use. In many computer vision, speech recognition and natural language processing tasks, deep-learning-based models have successfully produced state-of-the-art outcomes (Potamias et al. 2019; Koulierakis et al. 2020).