Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning
Estimating depth from a single image is a challenging problem, but it is also interesting due to the large amount of applications, such as underwater image dehazing. In this paper, a new perspective is provided; by taking advantage of the underwater haze that may provide a strong cue to the depth of...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-08-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/16/4580 |
| _version_ | 1797557928249524224 |
|---|---|
| author | Javier Pérez Mitch Bryson Stefan B. Williams Pedro J. Sanz |
| author_facet | Javier Pérez Mitch Bryson Stefan B. Williams Pedro J. Sanz |
| author_sort | Javier Pérez |
| collection | DOAJ |
| description | Estimating depth from a single image is a challenging problem, but it is also interesting due to the large amount of applications, such as underwater image dehazing. In this paper, a new perspective is provided; by taking advantage of the underwater haze that may provide a strong cue to the depth of the scene, a neural network can be used to estimate it. Using this approach the depthmap can be used in a dehazing method to enhance the image and recover original colors, offering a better input to image recognition algorithms and, thus, improving the robot performance during vision-based tasks such as object detection and characterization of the seafloor. Experiments are conducted on different datasets that cover a wide variety of textures and conditions, while using a dense stereo depthmap as ground truth for training, validation and testing. The results show that the neural network outperforms other alternatives, such as the dark channel prior methods and it is able to accurately estimate depth from a single image after a training stage with depth information. |
| first_indexed | 2024-03-10T17:23:10Z |
| format | Article |
| id | doaj.art-2aed74df6f534cb981b388f100788bd6 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T17:23:10Z |
| publishDate | 2020-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-2aed74df6f534cb981b388f100788bd62023-11-20T10:15:03ZengMDPI AGSensors1424-82202020-08-012016458010.3390/s20164580Recovering Depth from Still Images for Underwater Dehazing Using Deep LearningJavier Pérez0Mitch Bryson1Stefan B. Williams2Pedro J. Sanz3Department of Computer Science and Engineering, Jaume I University, Vicent Sos Baynat, s/n, 12071 Castellón, SpainAustralian Centre for Field Robotics, University of Sydney, Sydney, 2006 NSW, AustraliaAustralian Centre for Field Robotics, University of Sydney, Sydney, 2006 NSW, AustraliaDepartment of Computer Science and Engineering, Jaume I University, Vicent Sos Baynat, s/n, 12071 Castellón, SpainEstimating depth from a single image is a challenging problem, but it is also interesting due to the large amount of applications, such as underwater image dehazing. In this paper, a new perspective is provided; by taking advantage of the underwater haze that may provide a strong cue to the depth of the scene, a neural network can be used to estimate it. Using this approach the depthmap can be used in a dehazing method to enhance the image and recover original colors, offering a better input to image recognition algorithms and, thus, improving the robot performance during vision-based tasks such as object detection and characterization of the seafloor. Experiments are conducted on different datasets that cover a wide variety of textures and conditions, while using a dense stereo depthmap as ground truth for training, validation and testing. The results show that the neural network outperforms other alternatives, such as the dark channel prior methods and it is able to accurately estimate depth from a single image after a training stage with depth information.https://www.mdpi.com/1424-8220/20/16/4580deep learningdepth estimationunderwaterroboticsdehazingimage processing |
| spellingShingle | Javier Pérez Mitch Bryson Stefan B. Williams Pedro J. Sanz Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning Sensors deep learning depth estimation underwater robotics dehazing image processing |
| title | Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning |
| title_full | Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning |
| title_fullStr | Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning |
| title_full_unstemmed | Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning |
| title_short | Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning |
| title_sort | recovering depth from still images for underwater dehazing using deep learning |
| topic | deep learning depth estimation underwater robotics dehazing image processing |
| url | https://www.mdpi.com/1424-8220/20/16/4580 |
| work_keys_str_mv | AT javierperez recoveringdepthfromstillimagesforunderwaterdehazingusingdeeplearning AT mitchbryson recoveringdepthfromstillimagesforunderwaterdehazingusingdeeplearning AT stefanbwilliams recoveringdepthfromstillimagesforunderwaterdehazingusingdeeplearning AT pedrojsanz recoveringdepthfromstillimagesforunderwaterdehazingusingdeeplearning |