Adaptive Image Processing Methods for Outdoor Autonomous Vehicles
| Autor: | Lucie Halodova, Tomas Krajnik, Eliška Dvořáková, Jiří Ulrich, Tomas Vintr, Keerthy Kusumam, Filip Majer |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
0209 industrial biotechnology
Parameter control business.industry Computer science Feature extraction ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Navigation system Image processing 02 engineering and technology 020901 industrial engineering & automation Robustness (computer science) 0202 electrical engineering electronic engineering information engineering Learning methods 020201 artificial intelligence & image processing Computer vision Artificial intelligence business Relevant information Reflection mapping |
| Zdroj: | Modelling and Simulation for Autonomous Systems ISBN: 9783030149833 MESAS |
| DOI: | 10.1007/978-3-030-14984-0_34 |
| Popis: | This paper concerns adaptive image processing for visual teach-and-repeat navigation systems of autonomous vehicles operating outdoors. The robustness and the accuracy of these systems rely on their ability to extract relevant information from the on-board camera images, which is then used for the autonomous navigation and the map building. In this paper, we present methods that allow an image-based navigation system to adapt to a varying appearance of outdoor environments caused by dynamic illumination conditions and naturally occurring environment changes. In the performed experiments, we demonstrate that the adaptive and the learning methods for camera parameter control, image feature extraction and environment map refinement allow autonomous vehicles to operate in real, changing world for extended periods of time. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|