|
|
Analysis of Convolutional Neural Networks for Detection and Classification of Damages in Fingerprint Images
Fořtová, Kateřina ; Tinka, Jan (referee) ; Kanich, Ondřej (advisor)
The aim of this Master's thesis is to analyze detection and classification approaches using convolutional neural networks on the problem of fingerprint damage. The first part of the thesis deals with the study of literature related to biometrics and fingerprint processing with emphasis on possible diseases that may affect the fingertip area. Subsequently, the thesis focuses on neural network-based recognition. The thesis describes the architectures of convolutional neural networks and object detection approaches up to the latest research. Several detection methods for detection and classification of skin diseases affecting fingertip are proposed using modern architectures, different types of backbone networks and detection methods. Eight models based on four different detection and classification approaches are chosen for the experiments. Subsequently, each model is trained several times using configuration parameter adjustments. The models are assessed on the basis of various metrics and compared in terms of the use of the backbone network and the chosen method for detection. The best result of 76.875 % was achieved in the test of correctly detected and classified area on real fingerprint images. The most problematic disease for detection and classification was atopic eczema, whose symptoms can manifest in many ways.
|
|
|
Liveness Detection on Touchless Fingerprint Scanner
Fořtová, Kateřina ; Kanich, Ondřej (referee) ; Heidari, Mona (advisor)
This Bachelor's Thesis is focused on liveness detection of fingerprints with using touchless sensor. Work summarizes theoretical introduction to biometrics, fingerprint processing and some of present researches for liveness detection. The new approach is introduced with using Local Binary Pattern algorithm, Sobel and Laplacian operator and Wavelet transform. Artificial Neural Networks, Support Vector Machines and Decision Trees were used for final classification. Several experiments with dataset illuminated by lights with various wavelengths were realized. It was discovered, that fingerprints illuminated by red light reached the best accuracy 90.1% compared to other considered wavelenghts of visible light. The classification with vector based on Local Binary Pattern achieved average accuracy 89.8%, accuracy with vector based on Sobel and Laplacian operator was 91.5%. Several Wavelet families were used for Wavelet transform during experiments. The best accuracy achieved wavelets of Biorthogonal spline wavelet family (85.1%) and wavelets from Reverse biorthogonal spline wavelet family (86.6%).
|
|
|
Analysis of Convolutional Neural Networks for Detection and Classification of Damages in Fingerprint Images
Fořtová, Kateřina ; Tinka, Jan (referee) ; Kanich, Ondřej (advisor)
The aim of this Master's thesis is to analyze detection and classification approaches using convolutional neural networks on the problem of fingerprint damage. The first part of the thesis deals with the study of literature related to biometrics and fingerprint processing with emphasis on possible diseases that may affect the fingertip area. Subsequently, the thesis focuses on neural network-based recognition. The thesis describes the architectures of convolutional neural networks and object detection approaches up to the latest research. Several detection methods for detection and classification of skin diseases affecting fingertip are proposed using modern architectures, different types of backbone networks and detection methods. Eight models based on four different detection and classification approaches are chosen for the experiments. Subsequently, each model is trained several times using configuration parameter adjustments. The models are assessed on the basis of various metrics and compared in terms of the use of the backbone network and the chosen method for detection. The best result of 76.875 % was achieved in the test of correctly detected and classified area on real fingerprint images. The most problematic disease for detection and classification was atopic eczema, whose symptoms can manifest in many ways.
|
|
|
Liveness Detection on Touchless Fingerprint Scanner
Fořtová, Kateřina ; Kanich, Ondřej (referee) ; Heidari, Mona (advisor)
This Bachelor's Thesis is focused on liveness detection of fingerprints with using touchless sensor. Work summarizes theoretical introduction to biometrics, fingerprint processing and some of present researches for liveness detection. The new approach is introduced with using Local Binary Pattern algorithm, Sobel and Laplacian operator and Wavelet transform. Artificial Neural Networks, Support Vector Machines and Decision Trees were used for final classification. Several experiments with dataset illuminated by lights with various wavelengths were realized. It was discovered, that fingerprints illuminated by red light reached the best accuracy 90.1% compared to other considered wavelenghts of visible light. The classification with vector based on Local Binary Pattern achieved average accuracy 89.8%, accuracy with vector based on Sobel and Laplacian operator was 91.5%. Several Wavelet families were used for Wavelet transform during experiments. The best accuracy achieved wavelets of Biorthogonal spline wavelet family (85.1%) and wavelets from Reverse biorthogonal spline wavelet family (86.6%).
|
guest ::
RSS feed.