Authors: Adrian Burlacu, Andrei Baciu, Vasile Ion Manta and Simona Caraiman
in: System Theory, Control and Computing (ICSTCC), 2017 21st International Conference on
Abstract: Accurate ground area detection is one of the most important tasks in various stereo vision based applications, such as autonomous driving or assistive technologies for visually impaired. Correct assertion of the ground geometry improves obstacle detection algorithms by eliminating false positive locations in image. In this paper we provide an application oriented evaluation on the correlation of the ground geometry with the quality of the disparity map. The disparity map is processed in its V-map representations and two methods for 3D ground area points identification are discussed. Next, different types of surfaces are fitted and evaluated using real data from automotive and visually impaired assistive applications.
Ground-Geometry-Assessment-in-Complex-Stereo.pdf (498 downloads)