|
|
3D Display
Szkandera, Jan ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
This Bachelor's thesis goal is to design and realize system, which allows user to perceive 2D visual information as three-dimensional. 3D visual preception of 2D image is possible thanks to stereoprojection and projection dependecy of user location. The thesis mainly aims at secound of these problems.
|
|
|
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) ; 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.
|
|
|
Hand-Held 3D Scanner
Kukučka, Marek ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
The goal of this thesis is to implement a method of 3D reconstruction from pairs of images taken using a device that consists of two cameras. The method of stereo vision was chosen and the result is a sparse reconstruction of the scanned object. First, the cameras are calibrated and the acquired images are modified. In the next section, we look for corresponding key points in a pair of images. After obtaining the corresponding points, we an then perform their triangulation. For reconstruction from more then two consecutive images, we use recalculation of projection matrix. In this thesis an experiment is performed, with aim to test whether the reconstructed object corresponds in its dimensions to the real world.
|
|
|
Hand-Held 3D Scanner
Kukučka, Marek ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
The goal of this thesis is to implement a method of 3D reconstruction from pairs of images taken using a device that consists of two cameras. The method of stereo vision was chosen and the result is a sparse reconstruction of the scanned object. First, the cameras are calibrated and the acquired images are modified. In the next section, we look for corresponding key points in a pair of images. After obtaining the corresponding points, we an then perform their triangulation. For reconstruction from more then two consecutive images, we use recalculation of projection matrix. In this thesis an experiment is performed, with aim to test whether the reconstructed object corresponds in its dimensions to the real world.
|
|
|
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 Display
Szkandera, Jan ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
This Bachelor's thesis goal is to design and realize system, which allows user to perceive 2D visual information as three-dimensional. 3D visual preception of 2D image is possible thanks to stereoprojection and projection dependecy of user location. The thesis mainly aims at secound of these problems.
|
|
|
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.
|
|
| |
guest ::
