Header

UZH-Logo

Maintenance Infos

Learning-based Image Enhancement for Visual Odometry in Challenging HDR Environments


Gomez-Ojeda, Ruben; Zhang, Zichao; Gonzalez-Jimenez, Javier; Scaramuzza, Davide (2018). Learning-based Image Enhancement for Visual Odometry in Challenging HDR Environments. In: IEEE International Conference on Robotics and Automation (ICRA), 2018., Brisbane, 21 May 2018 - 25 May 2018, 1-8.

Abstract

One of the main open challenges in visual odometry (VO) is the robustness to difficult illumination conditions or high dynamic range (HDR) environments. The main difficulties in these situations come from both the limitations of the sensors and the inability to perform a successful tracking of interest points because of the bold assumptions in VO, such as brightness constancy. We address this problem from a deep learning perspective, for which we first fine-tune a deep neural network with the purpose of obtaining enhanced representations of the sequences for VO. Then, we demonstrate how the insertion of long short term memory allows us to obtain temporally consistent sequences, as the estimation depends on previous states. However, the use of very deep networks enlarges the computational burden of the VO framework; therefore, we also propose a convolutional neural network of reduced size capable of performing faster. Finally, we validate the enhanced representations by evaluating the sequences produced by the two architectures in several state-of-art VO algorithms, such as ORB-SLAM and DSO.

Abstract

One of the main open challenges in visual odometry (VO) is the robustness to difficult illumination conditions or high dynamic range (HDR) environments. The main difficulties in these situations come from both the limitations of the sensors and the inability to perform a successful tracking of interest points because of the bold assumptions in VO, such as brightness constancy. We address this problem from a deep learning perspective, for which we first fine-tune a deep neural network with the purpose of obtaining enhanced representations of the sequences for VO. Then, we demonstrate how the insertion of long short term memory allows us to obtain temporally consistent sequences, as the estimation depends on previous states. However, the use of very deep networks enlarges the computational burden of the VO framework; therefore, we also propose a convolutional neural network of reduced size capable of performing faster. Finally, we validate the enhanced representations by evaluating the sequences produced by the two architectures in several state-of-art VO algorithms, such as ORB-SLAM and DSO.

Statistics

Downloads

0 downloads since deposited on 21 Mar 2018
31 downloads since 12 months

Additional indexing

Item Type:Conference or Workshop Item (Paper), refereed, original work
Communities & Collections:03 Faculty of Economics > Department of Informatics
Dewey Decimal Classification:000 Computer science, knowledge & systems
Language:English
Event End Date:25 May 2018
Deposited On:21 Mar 2018 15:52
Last Modified:31 Jul 2018 05:56
Publisher:IEEE
OA Status:Green
Official URL:http://rpg.ifi.uzh.ch/docs/ICRA18_Gomez.pdf
Other Identification Number:merlin-id:16265

Download

Download PDF  'Learning-based Image Enhancement for Visual Odometry in Challenging HDR Environments'.
Preview
Content: Published Version
Filetype: PDF
Size: 1MB