|
|
Biometry based on iris images
Tobiášová, Nela ; Drahanský, Martin (referee) ; Kolář, Radim (advisor)
The biometric techniques are well known and widespread nowadays. In this context biometry means automated person recognition using anatomic features. This work uses the iris as the anatomic feature. Iris recognition is taken as the most promising technique of all because of its non-invasiveness and low error rate. The inventor of iris recognition is John G. Daugman. His work underlies almost all current public works of this technology. This final thesis is concerned with biometry based on iris images. The principles of biometric methods based on iris images are described in the first part. The first practical part of this work is aimed at the proposal and realization of two methods which localize the iris inner boundary. The third part presents the proposal and realization of iris image processing in order to classifying persons. The last chapter is focus on evaluation of experimental results and there are also compared our results with several well-known methods.
|
|
|
Detection and Recognition of Traffic Signs in Image
Spáčil, Pavel ; Hradiš, Michal (referee) ; Herout, Adam (advisor)
This work focuses on classification and recognition of traffic signs in image. It describes briefly some used methods a deeply describes chosen system including extensions and method for creating models needed for classification. There's described implementation of library and demonstration program including important pieces of knowledge discovered during development. There're also results of some experiments and possible enhancements in conclusion.
|
|
|
Handwritten Character Recognition Using Artificial Neural Networks
Smejkal, Vojtěch ; Fapšo, Michal (referee) ; Plchot, Oldřich (advisor)
Thesis deals with handwritten block letters and digits recognition using artificial neural networks. Text segmentation algorithms, feature extraction methods and backpropagation learning are explained. There are also described performed experiments with variety of configurations on datasets. Application with graphical user interface and interactive mouse-written text recognition was created to train new neural networks and test their effectivity.
|
|
|
Video Retrieval
Černý, Petr ; Mlích, Jozef (referee) ; Chmelař, Petr (advisor)
This thesis summarizes the information retrieval theory, the relational model basic and focuses on the data indexing in relational database systems. The thesis focuses on multimedia data searching. It includes description of automatic multimedia data content extraction and multimedia data indexing. Practical part discusses design and solution implementation for improving query effectivity for multidimensional vector similarity which describes multimedia data. Thesis final part discusses experiments with this solution.
|
|
|
Stress detection
Jindra, Jakub ; Vítek, Martin (referee) ; Němcová, Andrea (advisor)
Stress detection based on non-EEG physiological data can be useful for monitoring drivers, pilots, and also for monitoring of people in ordinary situation, where standard EEG monitoring is unsuitable. This work uses Non-EEG database freely available from Physionet. The database contains records of heart rate, saturation of blood oxygen, motion, a conductance of skin and temperature recorded for 3 type of stress alternated with relax state. Two final models were created in this thesis. First model for Binary classification stress/relax, second for classification of 4 different type of psychical state. Best results were reached using model created by decision tree algorithm with 8 features for binary classification and with 8 features for classification of 4 psychical state. Accuracy of final models is aproximately 95 % for binary model and 99 % for classification of 4 psychical state. All algorithms were implemented in Python.
|
|
|
Processing of iris images for biometric applications
Osičková, Kristýna ; Drahanský, Martin (referee) ; Kolář, Radim (advisor)
Biometrics is a method of recognizing the identity of a person based on unique biological characteristics that are unique to each person. The methods of biometric identification is currently becoming increasingly widespread in various sectors. This work is focused on the identification of a person by iris images. The introductory section describes the principles of the well-known methods for biometric applications and the next part describes the design method and its implementation in Matlab. In the practical part, fast radial symmetry method is used for detection of pupil, from which it derives further image processing. Two dimensional discrete welvet transform is used here. The proposed algorithm is tested on databases CASIA-Iris- Interval and database IITD.
|
|
|
Evolutionary Optimization of the EEG Classifier Feature Extractor
Ovesná, Anna ; Hurta, Martin (referee) ; Mrázek, Vojtěch (advisor)
This work focuses on the optimisation of EEG signal classification of alcoholics and control subjects using evolutionary algorithms with a multi-objective approach. The main goal is to maximise the accuracy, sensitivity and specificity of the classification algorithm and minimise the number of features used. Four different classifiers are used, namely Support Vector Machine, k-nearest neighbors, Naive Bayes and AdaBoost. The selection of the best features is optimised using three different evolutionary approaches, two of which convert multi-objective optimisation to single-objective using weighted summation or restricting the maximum number of features. The Pareto optimal solutions are found by the NSGA-II algorithm. Results show that the evolutionary algorithms, combined with appropriate classifiers, reliably distinguish a person with a tendency to alcoholism from one with a healthy relationship towards alcohol.
|
|
|
Stress detection
Jindra, Jakub ; Vítek, Martin (referee) ; Němcová, Andrea (advisor)
Stress detection based on non-EEG physiological data can be useful for monitoring drivers, pilots, and also for monitoring of people in ordinary situation, where standard EEG monitoring is unsuitable. This work uses Non-EEG database freely available from Physionet. The database contains records of heart rate, saturation of blood oxygen, motion, a conductance of skin and temperature recorded for 3 type of stress alternated with relax state. Two final models were created in this thesis. First model for Binary classification stress/relax, second for classification of 4 different type of psychical state. Best results were reached using model created by decision tree algorithm with 8 features for binary classification and with 8 features for classification of 4 psychical state. Accuracy of final models is aproximately 95 % for binary model and 99 % for classification of 4 psychical state. All algorithms were implemented in Python.
|
|
|
Handwritten Character Recognition Using Artificial Neural Networks
Smejkal, Vojtěch ; Fapšo, Michal (referee) ; Plchot, Oldřich (advisor)
Thesis deals with handwritten block letters and digits recognition using artificial neural networks. Text segmentation algorithms, feature extraction methods and backpropagation learning are explained. There are also described performed experiments with variety of configurations on datasets. Application with graphical user interface and interactive mouse-written text recognition was created to train new neural networks and test their effectivity.
|
|
|
Detection and Recognition of Traffic Signs in Image
Spáčil, Pavel ; Hradiš, Michal (referee) ; Herout, Adam (advisor)
This work focuses on classification and recognition of traffic signs in image. It describes briefly some used methods a deeply describes chosen system including extensions and method for creating models needed for classification. There's described implementation of library and demonstration program including important pieces of knowledge discovered during development. There're also results of some experiments and possible enhancements in conclusion.
|
guest ::
