Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning.

Sensors (Basel)

Department of Computer Science and Engineering, Jaume I University, Vicent Sos Baynat, s/n, 12071 Castellón, Spain.

Published: August 2020

AI Article Synopsis

  • Estimating depth from just one image poses challenges, especially in applications like underwater image dehazing, but it offers intriguing opportunities.
  • This paper introduces a neural network approach leveraging underwater haze as a depth cue to create depthmaps that enhance image quality and restore original colors, ultimately improving the accuracy of object detection in robotic systems.
  • Experiments on diverse datasets showed that the neural network significantly outperformed traditional methods, like dark channel prior techniques, in accurately estimating depth once trained with depth information.

Article Abstract

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.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7472610PMC
http://dx.doi.org/10.3390/s20164580DOI Listing

Publication Analysis

Top Keywords

depth single
8
single image
8
neural network
8
image
5
recovering depth
4
depth images
4
images underwater
4
underwater dehazing
4
dehazing deep
4
deep learning
4

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!