FACE AND FINGERPRINT PATTERN RETRIEVAL USING UNIQUE CODE

The use of number of biometric sources for human recognition is referred to as multibiometrics, which in turn mitigates limitations of single modal biometric systems by increasing identification accuracy, improving the population coverage, and imparting fault-tolerance and improving security. In this paper, we present a method for indexing multiple modality biometric databases which is completely based on index codes .And the index codes are generated by a inbuilt biometric matcher. The indexing mechanism which is executed individually for and the results are then combined together into a final list of all the potential candidates. The proposed indexing technique is based generation of index code depending upon the miniature position of orientation imagery for each modality and then by matching an input (probe) image against reference images which are stored in the database, which results in a set of, match scores. During identification, the index code which is generated for the input image is compared with the index codes of the identities which are already enrolled in the available database set so that find out a set of probable matches. The guide code of several number of modalities are fused together so that they can altogether percentage accuracy of indexing resulting in a and efficient and effective indexing system. This approach is based on a matcher, which is important part of every automated biometric identification system Because the generated index codes are very compact and their dissimilarity can be measured rapidly, this approach has a low memory storage requirements and can improve the overall response time of the system even for of the databases also