Sea-Thru: A Method for Removing Water From Underwater Images
By Derya Akkaynak, Tali Treibitz. Computer Vision and Pattern Recognition, 2019
489 citations recorded by the source corpus.
Robust recovery of lost colors in underwater images remains a challenging problem. We recently showed that this was partly due to the prevalent use of an atmospheric image formation model for underwater images. We proposed a physically accurate model that explicitly showed: 1)~the attenuation coefficient of the signal is not uniform across the scene but depends on object range and reflectance, 2)~the coefficient governing the increase in backscatter with distance differs from the signal attenuation coefficient. Here, we present a method that recovers color with the revised model using RGBD images. The \emph{Sea-thru} method first calculates backscatter using the darkest pixels in the image and their known range information. Then, it uses an estimate of the spatially varying illuminant to obtain the range-dependent attenuation coefficient. Using more than 1,100 images from two optically different water bodies, which we make available, we show that our method outperforms those using the atmospheric model. Consistent removal of water will open up large underwater datasets to powerful computer vision and machine learning algorithms, creating exciting opportunities for the future of underwater exploration and conservation.
Subfields: Computer Vision and Pattern Recognition, Media Technology
View Paper: https://www.semanticscholar.org/paper/192e3271d658b27b23c7cbbce178a74ccc1002db
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
In Media Technology
Citations in the corpus
- Underwater Image Enhancement via Medium Transmission-Guided Multi-Color Space Embedding (2021, methodology): “In addition to extra information [20], [21] and specialized hardware devices [22], underwater image enhancement can be roughly classified into two groups: traditional methods and deep learning-based methods.”
- Simultaneous Enhancement and Super-Resolution of Underwater Imagery for Improved Visual Perception (2020, background): “ty benefits for its robotic deployments. 2. Background Underwater image enhancement is an active research problem that deals with correcting optical image distortions to recover true pixel intensities [3,10]. Classical approaches use hand-crafted filters to improve local contrast and enforce color con”
- Underwater Image Enhancement With Hyper-Laplacian Reflectance Priors (2022, unknown): “In addition, the Sea-thru method [1] demanded the true depth map as an additional input to recover underwater color with the revised image formation model, and used spatially varying illuminant estimate to obtain the depth-dependent attenuation coefficient.”
- Underwater Single Image Color Restoration Using Haze-Lines and a New Quantitative Dataset (2018, methodology): “A method using this model [33] has shown superior results, however, it requires knowledge of the 3D scene structure.”
- Using labeled data to evaluate change detectors in a multivariate streaming environment (2009, unknown): “ICENet [26] and ICENetv2 [27] show that the two-branch network structure like BiSeNet [11] with a finer spatial branch and a deep context semantic branch is very suitable to segment the river ice with scale diversity characteristic.”