|
|
Fingerprint scanning
Kubiš, Michal ; Dobeš,, Michal (referee) ; Fedra, Petr (advisor)
Fingerprints are the oldest and most used form of biometric identification. A critical step is reliable extract minutiae from the fingerprint images. However fingerprint images are rarely of perfect quality, they may be degraded and corrupted due to natural variations in skin and sensing conditions. Thus, image enhancement techniques are necessary prior to minutiae extraction. This work includes implementation of three techniques for fingerprint image enhancement, minutiae extraction and consturction of fingerprint reading device. Experiments are realized with two sets of fingerprints to evaluate the performance of implemented techniques.
|
|
|
Processing of fingerprints
Kazík, Martin ; Hasmanda, Martin (referee) ; Říha, Kamil (advisor)
This diploma thesis deals with the identification of persons by their fingerprints. In the beginning, there is a brief overview of biometric methods. The issue of fingerprints and their automatic processing is very extensive and addresses a large amount of literature. This work describes the various stages processing fingerprints from receiving a fingerprint scan through the methods of enhancing fingerprint, binarization, the detection of minutiae to compare two fingerprints. After a short introduction of biometric techniques is briefly introduced the history of fingerprints as an instrument for identification of persons. Subsequently, there is the description of problems of obtaining fingerprints from older methods that require the transmission medium, to methods of obtaining fingerprints directly into a digital format. There are described basic principles of sensors and the principle of their function. Next part deals with methods of processing digital fingerprint images. These are methods for fingerprint image enhancement, methods for image thresholding, thinning of ridge lines, detection of global and local characteristics of the fingerprint. The next section introduces the basic methods of comparison and classification of fingerprints. The practical part describes the specific implementation of fingerprint processing chain. For each processing stage is implemented one selected method. The result is an application that allows displaying the output of each phase, the comparison of fingerprints 1:1 and also finds a fingerprint in the database, by comparing one fingerprint with many.
|
|
|
Processing of fingerprints
Kazík, Martin ; Hasmanda, Martin (referee) ; Říha, Kamil (advisor)
This diploma thesis deals with the identification of persons by their fingerprints. In the beginning, there is a brief overview of biometric methods. The issue of fingerprints and their automatic processing is very extensive and addresses a large amount of literature. This work describes the various stages processing fingerprints from receiving a fingerprint scan through the methods of enhancing fingerprint, binarization, the detection of minutiae to compare two fingerprints. After a short introduction of biometric techniques is briefly introduced the history of fingerprints as an instrument for identification of persons. Subsequently, there is the description of problems of obtaining fingerprints from older methods that require the transmission medium, to methods of obtaining fingerprints directly into a digital format. There are described basic principles of sensors and the principle of their function. Next part deals with methods of processing digital fingerprint images. These are methods for fingerprint image enhancement, methods for image thresholding, thinning of ridge lines, detection of global and local characteristics of the fingerprint. The next section introduces the basic methods of comparison and classification of fingerprints. The practical part describes the specific implementation of fingerprint processing chain. For each processing stage is implemented one selected method. The result is an application that allows displaying the output of each phase, the comparison of fingerprints 1:1 and also finds a fingerprint in the database, by comparing one fingerprint with many.
|
|
|
Fingerprint scanning
Kubiš, Michal ; Dobeš,, Michal (referee) ; Fedra, Petr (advisor)
Fingerprints are the oldest and most used form of biometric identification. A critical step is reliable extract minutiae from the fingerprint images. However fingerprint images are rarely of perfect quality, they may be degraded and corrupted due to natural variations in skin and sensing conditions. Thus, image enhancement techniques are necessary prior to minutiae extraction. This work includes implementation of three techniques for fingerprint image enhancement, minutiae extraction and consturction of fingerprint reading device. Experiments are realized with two sets of fingerprints to evaluate the performance of implemented techniques.
|
guest ::
