|
|
Detection of Red-Light Violation
Šorf, Milan ; Bartl, Vojtěch (referee) ; Špaňhel, Jakub (advisor)
This thesis is focused on automatic detection of red-light violation using static camera. Theoretical part of the thesis describes the principle of detection and introduces various methods of detecting traffic lights and cars as well as their reliability metrics. The thesis also illustrates implementation of the application for evaluation of these methods.
|
|
|
Reconstruction of Damaged Parts of Fingerprint Image
Halva, Vladislav ; Drahanský, Martin (referee) ; Kanich, Ondřej (advisor)
This bachelor thesis deals with the reconstruction of damaged fingerprint areas. The main goal is the design and implementation of an algorithm for the reconstruction of these regions. The designed algorithm consists of three main parts. The first part is fingerprint feature extraction and their derivation in damaged areas. The second part is a localization of damaged areas, which is based mainly on the structure of papillary lines. The last part is the damaged areas reconstruction itself. Gabor filter is used in this part of the process. The algorithm is implemented in C++ using the OpenCV library. An analysis of the reconstruction success rate is done afterwards. It is at first evaluated using the difference of quality between the input and processed fingerprint image, estimated by NIST NFIQ 2.0 and one other alternative tool for fingerprint image quality evaluation. The next step is a manual evaluation of the reconstruction success rate in various types of damaged areas.
|
|
| |
|
|
Automated Detection of Hate Speech and Offensive Language
Štajerová, Alžbeta ; Žmolíková, Kateřina (referee) ; Fajčík, Martin (advisor)
This thesis discusses hate speech and offensive language phenomenon, their respective definitions and their occurrence in natural language. It describes previously used methods of solving the detection. An evaluation of available data sets suitable for the problem of detection is provided. The thesis aims to provide additional methods of solving the detection of this issue and it compares the results of these methods. Five models were selected in total. Two of them are focused on feature extraction and the remaining three are neural network models. I have experimentally evaluated the success of the implemented models. The results of this thesis allow for comparison of the typical approaches with the methods leveraging the newest findings in terms of machine learning that are used for the classification of hate speech and offensive language.
|
|
|
Detection of closure of chamber doors for an electron scanning microscope
Obšitník, Michal ; Drexler, Petr (referee) ; Nešpor, Dušan (advisor)
The aim of this work is to design a suitable sensor for the detection of the chamber door position in an electron scanning microscope when the pressure in the vacuum chamber is changed and a control program is assembled to the selected sensor. The detected motion is within one millimeter. In the first part, the work is aimed at getting acquainted with the principle of functioning of the electron scanning microscope and with the basic types of sensors for position and mechanical force detection. Subsequently, the sensors are compared according to three basic categories: price, range, and resistance to external influences. Selected theory-based sensors are tested to exclude inappropriate sensors from use. After testing, the Interlink FSR-402 resistance sensor is selected. The scheme is connected to the sensor and the operation control program is created. The result of this work is a force sensing sensor, which is triggered by the pressure of chamber door on the chamber wall when creating a vacuum and a control program that controls the results measured from the sensor and adjusts them as required.
|
|
|
Thick-film sensors for heavy metals detection
Gajdoš, Libor ; Majzlíková, Petra (referee) ; Prášek, Jan (advisor)
This project describes the possibilities of heavy metals detections in solutions using differential pulse voltammetry with tree-electrode electrochemical thick-film sensors. In the theoretical part, the thick-film technology is described as well as thick-film pastes’ types and compositions, followed by introduction to electrochemistry and electro-analytical methods of measurement. This project also describes the three measured heavy metals – cadmium, lead and copper. In the practical part, heavy metals’ ions detection in solutions using carbon working electrode and determination of the detection limit and sensitivity are described and discussed. Finally the evaluation of obtained results and results comparison to hygienic limits are included.
|
|
|
DNS Service Attacks Simulation
Navrátil, Tomáš ; Kekely, Lukáš (referee) ; Kováčik, Michal (advisor)
The theme of this bachelor‘s thesis is the simulation and detection of cybernetic attacks on the DNS service. The goal was to simulate chosen attacks, analyze their behavior and create a tool capable of detecting these attacks in network traffic and replicating them for research purposes. The tool was able to successfully detect DNS attacks in normal network conditions. These results are discussed further at the conclusion of this paper, along with possible uses this application might have to other developers, and ways the program could be improved or extended in the future.
|
|
|
BitTorrent Traffic Detection
Florek, Daniel ; Hranický, Radek (referee) ; Polčák, Libor (advisor)
This thesis deals with a topic of BitTorrent protocol detection within a pcap file. I managed to design and implement a tool based on deep packet inspection which can detect IP adresses and their ports that were involved in a BitTorrent communication. This detection is extendable with flow analysis which may lead into more results but at the same time in a higher chance of false positives. Therefore this kind of detection is just optional.
|
|
|
Detection and mitigation of cyber attacks at local area networks
Racka, Jan ; Lieskovan, Tomáš (referee) ; Gerlich, Tomáš (advisor)
The bachelor thesis is focused on the detection and mitigation of flood attacks in local area networks. The thesis can be divided into two parts. In the theoretical part, first flooding attacks are described. Then, the problem of attack detection is discussed in depth, including the implemented detection methods. Subsequently, the classification of detection tools by location is discussed and examples of detection tools are given. The last theoretical section is devoted to network mapping methods and topology detection tools. In the practical part, the design of the IDS and the test network is discussed. The network consists of three end devices namely: the IDS, the victim and the attacker. A Mikrotik router is used to ensure connectivity between all the devices. The IDS has been implemented in Python and is composed of individual modules that extend its functionality. The most important module is the detection module, which contains detection methods against SYN Flood, UDP Flood, ICMP Flood attacks and one universal comprehensive method against all flood attacks. The ARP Scan module allowed the IDS to map the network and use ARP queries to detect the presence of endpoint devices in the network. The learning module made it easier to set up rules for each detection method by monitoring network traffic over a period of time. It then determines appropriate rule values from the detected data. The SSH module provided IDS with the ability to proactively respond to attacks and disconnect the attacker from the rest of the network. ARP Scans also use the SSH module to discover information about guests. The IDS has been tested in both virtual and real environments. The results show that the developed detection methods work and the IDS can stop the attack in a reasonable time. ARP Scanning was also tested and was able to detect new guests on average in the first pass. The effect of IDS on communication was also monitored and found to be minimal.
|
|
|
Automatic delineation of ECG signals
Jurek, Jakub ; Janoušek, Oto (referee) ; Vítek, Martin (advisor)
This project deals with basic description of ECG signal and some of known methods of delineation ECG´s individual parts. Next this work deals with detailed description of the method of authors Laguna, Jané, Caminal and realization of the complete delineation algorithm based on this method. Next this work deals with testing this algorithm on the CSE database, evaluation results and comparing results with results of authors of the method.
|
guest ::
