| Summary: | 碩士 === 國立臺北教育大學 === 資訊科學系碩士班 === 98 === Because of the advance of the technology and fast development of economy, the person wealth and security privacy have become important issues. People use key and password for security in general, but these kinds of tool are inconvenient to carry and maybe have the problem of forgetfulness and loss. Therefore, the personal identification with the biological characteristics is researched by professionals.
Biometrics recognitions is the use of a person’s inherent unique physical characteristics or behaviors such as face, fingerprint, palm, signature, iris, voice to verify their identity. In this thesis, we study and discuss the characteristics of fingerprint and utilize unique physical characteristics to verify their identity. Fingerprint is the most widely used characteristic for personal identification because the fingerprints have invariance and uniqueness. We measure the cost and performance that fingerprint is cheap and practically more than the others. Hence, fingerprint recognition has already been used in the court to verify criminal’s identity and it has become one of the important methods of people’s identity.
DSP(Digital Signal Processing) has a lot of advantages, such as high integration, reconfiguration, high speed, portable, high density, high capacity, low power consumption and infinitely reconfigurable. We propose an effective pre-process to enhance the fingerprint image. First, we obtain a fingerprint image from a sensor and then use then use Matlab search singular points for classification.
|
|---|
