Fingerprint Processing In The Resource-Constr-Ained Conditions

This thesis describes the fingerprint alignment algorithm and the suitable fingerprint pre-processing algorithm for resource-constrained applications in fingerprint verification system. The relative theory foundation, detailed procedures in realization, experimental results, conclusions and future work are included.In fingerprint alignment, we focus on the ridge curvature application. The ridge tracking algorithm, the curvature calculation algorithm, maximum curvature's computing and ICP algorithm are used. Experimental results and analysis are given at the same time.In the algorithm suitable for the resource-constrained applications, we aim to improve and optimize our existing fingerprint verification system. In this thesis, the system's optimization is divided into three parts: time, space and using the fixed-point number. All the details of analysis and realization are given in these three parts. Experimental results show that it can reduce the computational load without adversely affecting verification performance and can be used within a low cost, resource-constrained embedded environment in practical.