|
|
Detection of Weapons in 2D Image
Demčák, Ján ; Spurný, Martin (referee) ; Drahanský, Martin (advisor)
This bachelor thesis deals with detection of weapons in 2D image. In the theoretical part of the thesis the term weapon was defined and the possibilities of detection of weapons in image with using classic methods and deep neural networks were mentioned there. The key steps of image processing, objects classification and detection were described. The overview of frameworks, libraries was presented. To implement the pratical part of the thesis, 3 models were chosen. The first classic model with using HOG transformation. The second CNN model with priority target detection accuracy and with two different neural network architectures as classifiers. The third model with YOLO network architecture had as priority target real-time detection. The essential part of each model was choice, or more precisely creating suitable dataset. What followed was the construction and implementation of models and the evaluation of obtained data.
|
|
|
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, Localization and Determination of Chronic Wounds
Gulán, Filip ; Dvořák, Michal (referee) ; Drahanský, Martin (advisor)
The aim of this diploma thesis is to design and implement a multiplatform application for detection, localization and determination of the extent of chronic wounds. The application is intended to assist nurses, doctors and healthcare assistants to monitor and evaluate chronic wounds in the course of treatment. The application is based on the Typescript programming language, on the Ionic hybrid application framework and on the Electron desktop application framework. Chronic wound assessment runs on the server-side where the Python programming language is used. The Flask application framework is used for the RESTful application interface and the OpenCV library is used for image processing.
|
|
|
Detection and segmentation of lumbar vertebrae in 3D CT data
Nemček, Jakub ; Kolář, Radim (referee) ; Jakubíček, Roman (advisor)
This thesis deals with the detection and the segmentation of lumbar vertebrae in CT image datas. The described detection method is based on the use of a trained SVM classificator and histograms of oriented gradients as the image features. The detection method is applied on two-dimensional sagital slices of the CT image. The segmentation method is implemented as triangular mesh model deformation of models, that are obtained from averaged vertebrae in real CT datas. The first part of the thesis describes essential theoretical knowledge about the anatomy of the axial skeleton, computer tomography, image processing methods and about the detection and segmentation issues. The second part contains the algorithms realisation description, the evaluation and the discussion of the results. Applications of the algorithms in CAD systems is described at the end. The application of all of the points is done in the programming software Matlab.
|
|
|
Tabletop Object Detection
Timko, Martin ; Veľas, Martin (referee) ; Kapinus, Michal (advisor)
The aim of this thesis is to design and create a module for robotic platform PR2. This module has to detect objects in front of the robot on a table and enable various operations with these objects. This thesis describes methods of object detection which were used in the implementation of the module. The thesis also describes the methods used for design and implementation of this module. The module testing and evaluation of results are mentioned at the end of the thesis.
|
|
|
Static methods for detection DDoS attacks
Miško, Lukáš ; Dvořák, Jan (referee) ; Blažek, Petr (advisor)
This thesis contains a theoretical basic for solution to issue of network anomalies with use of static methods and it also contains software as a solution for detection of network attacks. The main point of thesis is detection of DoS (Denial of Service) attacks. In thesis is located an analysis of DoS attacks rate categorization. Further in thesis is located analysis of protocols TCP (Transmission Control Protocol) and UDP (User Datagram Protocol), their possible use to attacks SYN flood and UDP flood. Here are analysed three static methods and their detailed description. There is also a analysis of collected data and their comparison in the thesis. Thesis contains description and the results testing of software which is used to detect attacks in network, at the end.
|
|
|
I am here it is not
Daneková, Petra ; Hosnedlová, Klára (referee) ; Kohoutková, Karolína (advisor)
I have chosen the concept of mimicry, camouflage and disappearance because the subject of fusion and invisibility interests me. To vanish in the nature or in the city just with the help of body and clothes. For the purpose of this thesis I have created an outfit that will prove effectiveness of camouflage for the ordinary people watched by cameras everyday. By using this outfit I should find out whether digital devices are sophisticated enough to recognize a person even in the case that he has more legs or arms and his face is covered only with the eyes visible. In my thesis I am using several kinds of digital devices by which I am trying to figure out the effectiveness of created outfit. For comparation and better capture of person I have created two types of outfit. One of them is monochromatic while the second one, and more essential one, is printed with glitch. I am testing these two outfits both in exterior and interior. The outcome of the thesis are photographs and videos of person dressed in created outfits.
|
|
|
Verification of a New Canine Training Method for the Human Scent Identification
Haffner, Martin ; Pinc, Ludvík (advisor) ; Jaroslav, Jaroslav (referee)
Scent identification is a forensic method which is used in many countries. Principle of this method is always the same, but in each country are used different processes. This method is based on comparison of an odor collected at the crime scene and an odor collected from the suspect.
The aim of this study was to verify new methodology designed by Dr. Adee Schoon. This method consists of different dog training methodologies. In the commonly used method the dog gets an odor from his handler to sniff at before each particular line-up. This is match to sample method. That means the dog is able to compare two samples from the same person. In the modified methodology a dog once sniffs at the single odor placed on the ground. After distracting odors are added the dog is supposed to alert to the matching odor again. Thus the method under testing is more like based on the detection of target odor than on match-to-sample scenario.
Human odor samples were collected from palms of hands, metal tubes and the object belonging to the same person. Experimental persons were asked to wash their hands and let them to dry freely. At first, sorbent material Aratex was placed to palms of these persons for fifteen minutes and then it was closed and sealed in odorless glass jars. Then the metal tubes were given to hold to experimental persons for five minutes. After this time metal tubes were placed into glass jars with odorless Aratex which was later used as scent samples. The metal tubes were removed from glass jars after thirty minutes. Scent samples from objects were collected similarly as from metal tubes, these samples served as a corpus delict. The starting scent samples and target scent samples were collected by a different persons. Additional odors were collected using the same protocol and under similar conditions, so none of the odor samples in a line-up were more attractive for the dog than the others.
For the training three years old female belgian shepherd malinois was used, trained by the author of the project. Intrinsic matching procedure was always three times repeated. At first, the dog sniffed at the scent sample from the palms of hands. The line-up was arranged of scent samples collected from metal tubes and objects. One of these scent samples was the target scent (metal tubes). After comparison of metal tubes scents, the target scent was replaced by a scent sample collected from corpus delicti. In case the dog correctly indicated target scent, the result was recorded as correct.
For statistical evaluation Bernoulli distribution was used (P < 0.01). Over the whole experiment only one dog was used, and so it cannot be concluded that this method is easier for the dog than the traditional one, however the study demonstrated that such a method is basically usable as a tool by which dogs can be trained to identify individual human scents. Fischer´s test did not show any differences between comparisons based on the type of an object.
|
|
|
Recognition of Skin Diseases on Fingers of a Human Hand
Šesták, Martin ; Orság, Filip (referee) ; Drahanský, Martin (advisor)
This bachelor thesis deals with the design of methods and application for recognition of the fingerprints from diseased and healthy fingers and the subsequent detection of selected diseases. The first part describes basic principles of biometrics, introduction to fingerprints and their processing, and introduction to the topic of fingerprints affected by skin diseases. The next section describes the implementation of the application and its results. The application was tested on a database of research group STRaDe, from the Department of Intelligent Systems, Faculty of Information Technology, Brno University of Technology, which contains approximately 380 test fingerprints.
|
|
|
Traffic Signs Detection and Localisation
Kudláč, Ondrej ; Španěl, Michal (referee) ; Veľas, Martin (advisor)
This thesis aims to design the traffic signs detection and localization system using RGB image and 3D LiDAR data leveraging the the existing solutions. Traffic sign detection is based on the shape analysis. Then, the LIDAR data are used for the localization of previously detected signs. The created solution consists of two main components: the detector and locator, each able to operate independently.
|
guest ::
