Biometric authentication is a system which is capable to identifying a (01) based on the inherent physical or behavioral traits associated with that person. In recent years the application of biometrics in authentication systems, increase gradually, one of the main reasons for its popularity is that biometric traits like a fingerprint, face and iris feature of a person are (02) during the whole life time and it is not easily guessed, forgotten and misplaced. Biometric based system is more secure and accurate compared to the traditional system which is based on user personal identification number (PIN) and the user set PIN can be easily guessed by the third party. Unimodal system which is based on single modality has several inherent problems like intra-class variation, spoofing attacks and failure-to-enroll rate. To overcome this limitation multibiometric is a good option where we can use (03) one modality at a time to improve the performance and accuracy level of the systems. A fusion process plays an important role in multibiometric where the features of more than one modality are combined together. The whole fusion process can be classified as sensor level, feature level (combining features from different biometrics), score level (combining the genuine and imposter score), decision level (combining the decisions) and rank level (combining the ranks). Among all the fusion methods score level fusion is very popular and simple; lots of research has been done in the area of score level fusion. Researchers prefer to use score level fusion due to ease of combining matching score. Further score level fusion technique can be categorized as a) transformation-based score fusion like sum rule, weighted sum rule and the product rule. b) Classifier based like support vector machine (SVM) and c) density based fusion like likelihood ratio test. It is very (04) to deal with the feature level fusion because at this level the feature sets generated from multiple modalities are different in nature as in the case of fusion of fingerprint and face. Researchers preferred (05) use score level fusion due to ease of combining matching score. Due to lack of information it is very difficult to select a decision level, as a fusion strategy in a multibiometric system. In feature level fusion due to the large dimensionality of the feature set, it may affect the performance of the system, so some feature selection technique is used to identify and remove the irrelevant and redundant information and may allow learning algorithms to operate faster and more effectively.
（6）
A．person
B．server
C．device
D．computer
（7）
A．diferernt
B．safe
C．same
D．complete
（8）
A．of
B．more than
C．on
D．a(chǎn)s
（9）
A．difficult
B．easy
C．relaxed
D．fast
（10）
A．with
B．to
C．for
D．by