|
|
Automatic Photography Categorization
Veľas, Martin ; Řezníček, Ivo (referee) ; Španěl, Michal (advisor)
This thesis deals with content based automatic photo categorization. The aim of the work is to experiment with advanced techniques of image represenatation and to create a classifier which is able to process large image dataset with sufficient accuracy and computation speed. A traditional solution based on using visual codebooks is enhanced by computing color features, soft assignment of visual words to extracted feature vectors, usage of image segmentation in process of visual codebook creation and dividing picture into cells. These cells are processed separately. Linear SVM classifier with explicit data embeding is used for its efficiency. Finally, results of experiments with above mentioned techniques of the image categorization are discussed.
|
|
|
3D Mapping from Sparse LiDAR Data
Veľas, Martin ; Hofierka,, Jaroslav (referee) ; Kaartinen,, Harri (referee) ; Herout, Adam (advisor)
Tato práce se zabývá návrhem nových algoritmů pro zpracování řídkých 3D dat senzorů LiDAR, včetně kompletního návrhu batohovího mobilního mapovacího řešení. Tento výzkum byl motivován potřebou takových řešení v oblasti geodézie, mobilního průzkumu a výstavby. Nejprve je prezentován iterační algoritmus pro spolehlivou registraci mračen bodů a odhad odometrie z měření 3D LiDARu. Problém řídkosti a velikosti těchto dat je řešen pomocí náhodného vzorkování pomocí Collar Line Segments (CLS). Vyhodnocení na standardní datové sadě KITTI ukázalo vynikající přesnost oproti známému algoritmu General ICP. Konvoluční neuronové sítě hrají důležitou roli ve druhé metodě odhadu odometrie, která zpracovává kódovaná data LiDARu do 2D matic. Metoda je schopna online výkonu, zatímco je zachována přesnost, když požadujeme pouze parametry posunu. To může být užitečné v situacích, kdy je vyžadován online náhled mapování a parametry rotace mohou být spolehlivě poskytnuty např. senzorem IMU. Na základě algoritmu CLS bylo navrženo a implementováno batohové mobilní mapovací řešení 4RECON. S využitím kalibrovaného a synchronizovaného páru LiDARů Velodyne a s nasazením řešení GNSS/INS s duální anténou, byl vyvinut univerzální systém poskytující přesné 3D modelování malých vnitřních i velkých otevřených prostředí. Naše hodnocení prokázalo, že požadavky stanovené pro tento systém byly splněny -- relativní přesnost do $5$~cm a průměrná chyba georeferencí pod $12$~cm. Poslední stránky obsahují popis a vyhodnocení další metody založené na konvolučních neuronových sítích -- navržených pro segmentaci země v mračnech bodů 3D LiDARu. Tato metoda překonala současný stav techniky v této oblasti a představuje způsob, jakým může být sémantická informace vložena do 3D laserových dat.
|
|
|
Mobile App for Sharing Information about Presence in a Location
Vlk, Tomáš ; Veľas, Martin (referee) ; Herout, Adam (advisor)
The aim of this work is to create a user-friendly mobile application for Android that will make it easier to share ones's location while keeping their privacy. Instead of the GPS coordinates of the current location, the application only shares the information about the user's presence at the place. For easier sharing, it also offers automatic detection of the presence at a specific place. The result of this work is a application with a user interface that respects the Material Design, which makes it possible to determine the device's presence at a place with a radius greater than 50 meters, but there is no active polling. The application which has been created does not need to be run at all (or any of its background services) and it can detect the device's presence at a specific place and receive messages from other devices. The biggest advantage of this approach is very low energy consumption in contrast to other solutions.
|
|
|
Vehicle for Small and Remote Space Mapping
Koupý, Pavel ; Veľas, Martin (referee) ; Beran, Vítězslav (advisor)
This work describes process of making a robotic vehicle for remote mapping of small indoor spaces and areas. Such process involves design and construction, selection and connection of used electronic parts and implementation of interface witch wil provide set of tools as remote control, camera stream or autonomous space search.
|
|
|
Localisation of Mobile Robot in the Environment
Němec, Lukáš ; Hradiš, Michal (referee) ; Veľas, Martin (advisor)
This paper addresses the problem of mobile robot localization based on current 2D and 3D data and previous records. Focusing on practical loop detection in the trajectory of a robot. The objective of this work was to evaluate current methods of image processing and depth data for issues of localization in environment. This work uses Bag of Words for 2D data and environment of point cloud with Viewpoint Feature Histogram for 3D data. Designed system was implemented and evaluated.
|
|
|
Searching for Similar 3D Models
Šťáva, Zdeněk ; Veľas, Martin (referee) ; Španěl, Michal (advisor)
This paper deals with searching similar 3D models in a database containing up to thousands of models. It focuses in particular on the comparison of existing descriptors used to describe 3D models and the subsequent evaluation of similarity between models. In particular, the descriptors Rotation Invariant Spherical Harmonics and 3D Zernike Descriptor are compared. It also describes the use of libraries to extract these descriptors and to design of various experiments with these libraries over several object databases. It examines the effect of scale, translation, deformation and rotation of different 3D models on the resulting descriptor and the overall accuracy of both selected methods. These results compare.
|
|
|
Automatic People Counting from Panoramic Photography
Blucha, Ondřej ; Kolář, Martin (referee) ; Veľas, Martin (advisor)
This bachelor thesis deals with automatic people counting from panoramic photography. This is very useful for counting large number of people, such as on the stadium or on the concerts. It consists of the two parts. The first one is image stitching, which process the images by the feature-based methods. The second part is people counting using face detection, where were used Viola-Jones detector. The ideal setting of parameters for used methods was experimentally selected.
|
|
|
Visual Car-Detection on the Parking Lots Using Deep Neural Networks
Stránský, Václav ; Veľas, Martin (referee) ; Rozman, Jaroslav (advisor)
The concept of smart cities is inherently connected with efficient parking solutions based on the knowledge of individual parking space occupancy. The subject of this paper is the design and implementation of a robust system for analyzing parking space occupancy from a multi-camera system with the possibility of visual overlap between cameras. The system is designed and implemented in Robot Operating System (ROS) and its core consists of two separate classifiers. The more successful, however, a slower option is detection by a deep neural network. A quick interaction is provided by a less accurate classifier of movement with a background model. The system is capable of working in real time on a graphic card as well as on a processor. The success rate of the system on a testing data set from real operation exceeds 95 %.
|
|
|
Cloud Solution for 3D Models Processing
Klemens, Jakub ; Veľas, Martin (referee) ; Španěl, Michal (advisor)
This thesis deals with the possibilities of processing 3D models in cloud applications. A C++ library called Cloud3D has been designed and implemented. The resulting library is used to quickly create client-server applications. The library is divided into three separate parts: Client, Service Provider and Load Balancer. The service provider runs in the cloud and provides 3D model processing services to client applications. The biggest advantage of Cloud3D is the ease of creating new applications with its help. Other benefits include scalability, assured implementation of look-a-side Load Balancer, and security ensured by the use of SSL certification.
|
|
|
Road Detection for Autonomous Car
Komora, Matúš ; Veľas, Martin (referee) ; Španěl, Michal (advisor)
his thesis deals with detection of the road adjacent to an autonomous vehicle. The road is recognition is based on the Velodyne LiDAR laser radar data. An existing solution is used and extended by machine learning - a Support Vector Machine with online learning. The thesis evaluates the existing solution and the new one using a KITTI dataset. The reliability of the road recognition is then computed using F-measure.
|
guest ::
RSS feed.