|
|
3D Scene Reconstruction from Images
Ambrož, Ondřej ; Řezníček, Ivo (referee) ; Španěl, Michal (advisor)
Existing systems of scene reconstruction and theorethical basics necessary for scene reconstruction from images data are described in this work. System of scene reconstruction from video was designed and implemented. Its results were analyzed and possible future work was proposed. OpenCV, ArtToolKit and SIFT libraries which were used in this project are also described.
|
|
|
3D model
Sládeček, Martin ; Petyovský, Petr (referee) ; Richter, Miloslav (advisor)
This paper concerns the task three-dimensional scene using image sequence obtained with an ordinry camera. First portion of this thesis outlines the principles used in solving of the task, the second chapter describes a reconstucion algorithm and it's implementation in the Python programming language. The output of this program is demonstrated on several selected scenes. Final remarks discuss the quality of resulting models, shortcomings of the program and possible improvements.
|
|
|
Detection and Correspondence of Image Features
Hasmanda, Martin ; Kohoutek, Michal (referee) ; Říha, Kamil (advisor)
The main goal of this bachelor‘s thesis was obtain basic knowledge about image processing. Especially was this work specialized on features detection in images captured from different perspectives and for finding correspondences between those images. Preliminary were to be described the basic principles for understanding computer vision such as perspective projection, description model of the camera and two views geometry. From methods of the detection was introduced best-known and most widely used of the detectors Harris corner detector. He is independent of images rotation and he is analyzed in detail. Further was described SIFT detector, that is independent of images scale. In this work further be described to several methods for finding correspondences of images features. First were to be described and derived two basic transformation matrixes that arrange to the association with features of two images. The first homography matrix describes transformation of two 2D views and fundamental matrix. Fundamental matrix obtains from two images full information of captured 3D scene and projection matrixes of cameras. For to primary definition correspondences were to be used to methods SSD and NCC. These methods match correspondences after similarities surroundings of features. These methods unfiled correct assignment features. Therefore uses stochastic RANSAC algorithm. The RANSAC algorithm was in detail described in this work in basic form and further modified on MLESAC algorithm. This algorithm can find better correspondences than RANSAC. In the end was described simple application for implementation introduced methods.
|
|
|
3D Reconstruction from two views
Myška, Milan ; Minář, Jiří (referee) ; Hasmanda, Martin (advisor)
This bachelor thesis deals with 3D reconstruction process using two 2D pictures of one scene taken with uncalibrated cameras. The first chapter focuses on mathematical theory behind this process. Epipolar geometry, locating keypoints method SURF, fundamental matrix, stereo rectification and stereo correspondence using block-matching method are explained here. The next chapter of my thesis describes a practical view of this topic and my own C++ program is introduced at the same time.
|
|
|
3D Reconstruction from two views
Myška, Milan ; Minář, Jiří (referee) ; Říha, Kamil (advisor)
This bachelor thesis deals with 3D reconstruction process using two 2D pictures of one scene taken with uncalibrated cameras. The first chapter focuses on mathematical theory behind this process. Epipolar geometry, locating keypoints method SURF, fundamental and essential matrix, projection matrices, stereo rectification process and searching for stereo correspondences are explained here. The next chapter of my thesis describes a practical view of this topic and my own C++ program is introduced at the same time.
|
|
|
3D model from multiview
Zemčík, Tomáš ; Klečka, Jan (referee) ; Richter, Miloslav (advisor)
Reconstruction of 3D model from multiple view images allows us to perform a contactless measurement of real world objects. This work contains theoretical and mathematical background necessary for understanding problems of 3D reconstruction. It also proposes possible hardware and software solutions for the task of measuring a hood with EEG sensors while on patients head.
|
|
| |
|
|
Processing of Stereoscopic Video Sequence
Hasmanda, Martin ; Šmirg, Ondřej (referee) ; Říha, Kamil (advisor)
The main goal of this master’s thesis was get up used methods for observation the stereoscopic scene with one couple of cameras and find out good solving for processing these resulting pictures for two-view and multiple-view autostereoscopic displays for three-dimensional perception. For methods for acquisition video was introduced two methods. They were method “Off-axis” with parallel camera axis and method “Toe in” with intersections axis. For fit method was choice the method named as “Off-axis“. It was not produces the vertical parallax and in detail was in this work described principle of this method. Further were describe principles off used methods for three-dimensional perception namely from the oldest method named anaglyph after methods for viewing at autostereoscopic displays. The Autostereoscopic displays were main goal of this thesis and so their principles were described in details. For production the result image for autostereoscopic displays was used generation intermediate images between left and right camera. Resulting videos were acquisition for testing scene in created in 3D studio Blender, where was possible setting system of cameras exactly parallel axis. Then were introduce principles processing video where was extract from the couple of cameras where were connected to PC for help digitizing card and next time with two web cameras. Here is not guaranteed exact parallel axis system. Therefore this work try for real cameras achieve exactly parallel axis system by the help of transformations of frames with stereo rectification. Stereo rectification was solving with OpenCV libraries and was used two methods. Both methods work from principles epipolar geometry that was described in this work also in detail. First method rectifies pictures on the basis fundamental matrix and found correspondences points in two images of the scene and second method rectifies pictures from knowledge intrinsic and extrinsic parameters of stereoscopic system of cameras. In the end of this work was described application for implementation introduced methods.
|
|
|
Automatic calculation of relative orientation of two cameras
Adámek, Daniel ; Nováček, Petr (referee) ; Klečka, Jan (advisor)
To determine a relative orientation of two cameras, it is necessary to understand wide range of fields. Therefore, this thesis showes principles of basic methods of detection and description of keypoints in an image as well as methods of matching these keypoints. The paper also provides brief summary of epipolar geometry and linear triangulation. C++ implementation of presented theory was proposed and its disadvantages were discussed.
|
|
|
Automatic Processing of Recordings from In-Car Camera
Klos, Dominik ; Herout, Adam (referee) ; Španěl, Michal (advisor)
This thesis deals with the automatic processing of recordings from in-car camera, which is used for estimating distances in a scene filmed. To achieve this goal, it uses the methods of image processing, spatial geometry of the scene and 3D reconstruction, whose final output is a partial model. For this aim, the Bundler tool is used. Results of reconstruction are visualized in the application, which then allows to estimate the distance of objects from the camera or the distance between each other.
|
guest ::
