|
Abstract:
|
Multibiometrics refers to the process of consolidating the evidence presented by multiple biometric sources in order to enhance the recognition accuracy of a biometric system. Thus, multibiometric systems combine the information presented by multiple biometric sensors (e.g., thermal/IR cameras), algorithms (e.g., minutiae-/ridge-based fingerprint matchers), samples (e.g., frontal-/side-profile fae images), units (e.g., left/right irises) or traits (e.g., face/fingerprint). This work examines the various levels of fusion as well as the different fusion techniques that have been described in the literature. The impact of correlated biometric sources on the performance of different fusion rules at the score level is analyzed. Experimental results help in understanding the role of correlation, diversity and imbalance between biometric sources on the performance of different fusion rules.
|
