|
|
Stereo Reconstruction with Deep Neural Networks
Letanec, Richard ; Herout, Adam (referee) ; Španěl, Michal (advisor)
The aim of this thesis is to design and train a neural network model capable of estimating a disparity map from a pair of images. It will then be possible to create a depth map and point cloud from the estimated disparity map. Such a process is called stereo reconstruction. Solving this task consists of two steps -- choosing a suitable dataset and choosing a suitable neural network architecture. In my work, I compared two neural network architectures that I trained on the DrivingStereo dataset, consisting of paired images photographed from the roof of a car, and retrained and evaluated on the KITTI 2015 dataset, consisting of images of the same type. As the first neural network architecture, I chose ES-Net, which uses an approach based on a sequence of residual blocks and convolutional layers. As the second architecture, I chose CREStereo, which uses an iterative approach based on recurrent layers to predict the disparity map. In all benchmark tests, the CREStereo architecture achieves better accuracy.
|
|
|
Stereo reconstruction of vehicles' cross section
Boch, Jan ; Richter, Miloslav (referee) ; Honec, Peter (advisor)
This master thesis deals with the design of a stereo system for processing vehicle images and subsequent reconstruction of their 3D model. The aim of the thesis is to reconstruct the vehicle model to control its dimensions and load on which restrictions are imposed in case of the use of critical transport infrastructure (tunnels, bridges). The thesis begins with a mathematical description of scene capture and continues with a research of theoretical possibilities of system configuration. One Intel RealSense 3D camera and one RGB camera were chosen as the most suitable configuration. The software solution works with the design of several different algorithms of which it is worth mentioning reconstruction using depth maps or triangulation.The resulting models in the last chapter contain data only from the aforementioned 3D camera. The model from the 3D camera could not be improved with the images from the RGB camera and camera can therefore be declared redundant. Resulting 3D model is still sufficient for the purpose of vehicle inspection and in case of continued work, another, more optimal configuration should be designed.
|
|
|
Image Completion Using Depth Maps
Valeš, Ondřej ; Brejcha, Jan (referee) ; Čadík, Martin (advisor)
The aim of this thesis is to design and implement aplication for dataset driven scene completion utilizing data from similar dataset images and test the posibility of generating data used in reconstruction directly from depth map. For scene matching in dataset novel modification of GIST descriptor including depth data is used. Furthermore, depth maps can be used to split reconstructed image into parts with similar depth, simplifying reconstruction process. Part of this thesis is also computing GIST descriptors for datset images and implementation of tools for searching datset for similar images using depth maps.
|
|
|
Calculating disparity map from color stereo images
Kulíková, Barbora ; Nováček, Petr (referee) ; Klečka, Jan (advisor)
This bachelor’s thesis deals with a creation of the depth maps. The first chapter concerns with the physiology of a human space perception and the methods of displaying the 3D content which are the topics closely related to the depth map, its creation and practical usage. Subsequently, there is a chapter focused on the description of the used methods of the image processing. The fundamental theoretical chapter deals with the methods of computing the disparity and used principles. In the practical part of the thesis an application has been made with a user interface in the Matlab environment. The application enables the user to create the disparity maps through the local and global methods. The functionalism of the application and the implementation of the methods are experimentaly verified. An experiment comparing the methods and analyzing influence of the local method parameters on the quality of the depth map has been made. The last part of the thesis was to create a simple stereopicture database.
|
|
|
Virtual Gate for Counting the Passing of Persons
Chudý, Andrej Oliver ; Dvořák, Michal (referee) ; Drahanský, Martin (advisor)
The goal of this thesis is to propose a viable technique of detecting people passing through a virtual line (any opening without a turnstile or other form of the physical obstacle). Two software solutions were implemented to achieve high accuracy. First one processes 2D data output from webcam and the other one processes depth data from the 3D imaging sensor. Both software solutions were deployed and tested in a pharmaceutical storage room, where they achieved the accuracy of 98%. The software proposed in this thesis is thus reliable enough to become the basis of access control and security systems or for real-time evaluation of visitor rate statistics of commercial properties and events. The thesis also contains a comparative analysis of 3 widely used depth sensors.
|
|
|
Dynamic Image Presentations Using Depth Maps
Hanzlíček, Jiří ; Brejcha, Jan (referee) ; Čadík, Martin (advisor)
This master's thesis focuses on the dynamic presentation of still photography using a depth map. This text presents an algorithm that describes the process of creating a spatial model which is used to render input photography so that the movement of virtual camera creates parallax effect due to depth in image. The thesis also presents an approach how to infill the missing data in the model. It is suggested that a guided texture synthesis is used for this problem by using rendering outputs of the model themselves as guides. Additional information in model allows the virtual camera to move more freely. The final result of the camera movement can be saved to simple video sequence which can be used for presenting the input photography.
|
|
|
Image processing and depth maps to create an interactive model
Gelo, Lukáš ; Adámek, Roman (referee) ; Appel, Martin (advisor)
This bachelor’s thesis deals with creation of communication protocol for Intel® RealSense™ cameras in MATLAB, creation of program which finds certain objects in image using depth map and implementation of this program into mechatronic exponate. Theoretical part deals with methods of depth and distance scanning. It clarifies principles of contactless and nondestructive measurement of depth and distance along with comparison of their advantages and disadvantages.
|
|
|
Focus techniques of optical measurement of 3D features
Macháček, Jan ; Honec, Peter (referee) ; Janáková, Ilona (advisor)
This thesis deals with optical distance measurement and 3D scene measurement using focusing techniques with focus on confocal microscopy, depth from focus and depth from defocus. Theoretical part of the thesis is about different approaches to depth map generation and also about micro image defocusing technique for measuring refractive index of transparent materials. Then the camera calibration for focused techniques is described. In the next part of the thesis is described experimentally verification of depth from focus and depth from defocus techniques. For the first technique are shown results of depth map generation and for the second technique is shown comparison between measured distance values and real distance values. Finally, the discussed techniques are compared and evaluated.
|
|
|
Design of a New Method for Stereovision
Kopečný, Josef ; Drahanský, Martin (referee) ; Orság, Filip (advisor)
This thesis covers with the problems of photogrammetry. It describes the instruments, theoretical background and procedures of acquiring, preprocessing, segmentation of input images and of the depth map calculating. The main content of this thesis is the description of the new method of stereovision. Its algorithm, implementation and evaluation of experiments. The covered method belongs to correlation based methods. The main emphasis lies in the segmentation, which supports the depth map calculation.
|
|
|
Robotic Shopping Cart
Krzeminský, Juraj ; Kapinus, Michal (referee) ; Materna, Zdeněk (advisor)
This Bachelor thesis deals with design and implementation of control unit for robotic shopping cart. Its main task is to follow a chosen person through a mall and serve as bag carriage. The text describes basic principles of human detection and tracking in images and also the method of capturing them with RGB-D camera kinect. Rest of this text contain detailed description of created application and implementation of used methods and algorithms.
|
guest ::
