Underwater Datasets

English | 简体中文

This repository is mainly used to organize the datasets of underwater scenes, including the existing open source underwater datasets as detailed as possible. If you know that the open source underwater datasets are not included in our repository, you can share them if you are willing to Submit an issue for us to add the datasets, or contact us via email:

@xinzhichao xinzhichao@stu.ouc.edu.cn. [Github]

@zhengziqiang zhengziqiang1@gmail.com. [Github]

---

---

1.Image Enhancement

The underwater scene dataset in this part does not contain the depth map corresponding to the image

• EUVP dataset

  • EUVP dataset： EUVP dataset
  • Paper： Fast Underwater Image Enhancement for Improved Visual Perception
  • Code：Github
  • Dataset introduction:

• underwater_imagenet dataset

  • Underwater_imagenet dataset： Underwater_imagenet
  • Paper： Enhancing Underwater Imagery using Generative Adversarial Networks(UGAN)
  • Code：Github
  • Dataset introduction:

• UIEBD dataset

  • UIEBD dataset
    • UIEB Raws
    • UIEB References
  • Paper： An Underwater Image Enhancement Benchmark Dataset and Beyond(Water-Net)
  • Homepage：Homepage
  • Dataset introduction:For detailed dataset introduction, please check the author's official Homepage. It should be noted that this dataset is used for underwater image restoration, so two data sets are provided in pairs, one is raw and the other is the corresponding label data set references. .

• SUIM dataset

  • SUIM dataset SUIM dataset
  • Paper： Semantic Segmentation of Underwater Imagery: Dataset and Benchmark
  • Homepage：Homepage
  • Dataset introduction: This dataset is an underwater segmentation dataset, which contains already marked segmentation tags. For a detailed introduction of the dataset, please see the detailed introduction on the Homepage.

• RUIE benchmark dataset

  • RUIE benchmark dataset RUIE benchmark dataset
  • Paper： Real-world Underwater Enhancement: Challenges, Benchmarks, and Solutions(RUIE-Net)
  • Dataset introduction:

• U-45 dataset

  • U-45 dataset U-45 dataset
  • Paper： A Fusion Adversarial Underwater Image Enhancement Network with a Public Test Dataset
  • Dataset introduction:

• Water-GAN dataset

  • Water-GAN dataset Water-GAN dataset
  • Paper： WaterGAN: Unsupervised Generative Network to Enable Real-time Color Correction of Monocular Underwater Images
  • Code：Github
  • Dataset introduction:

• SQUID dataset

  • SQUID dataset SQUID dataset
  • Paper： Underwater Single Image Color Restoration Using Haze-Lines and a New Quantitative Dataset
  • Code：Github
  • Dataset introduction:

• Virtual periscope dataset

  • Virtual periscope dataset Virtual periscope dataset
  • Paper： Triangulation in Random Refractive Distortions
  • Dataset introduction:

• Color correction dataset

  • Color correction dataset Color correction dataset
  • Dataset introduction:

• Color restoration dataset

  • Data: Color restoration dataset
  • Paper： Underwater Single Image Color Restoration Using Haze-Lines and a New Quantitative Datasett
  • Code：Github
  • Dataset introduction:

• TURBID dataset

  • Data: TURBID dataset
  • Paper： A dataset to evaluate underwater image restoration methods
  • Dataset introduction:

• OceanDark dataset

  • Data: OceanDark dataset
  • Paper： A Contrast-Guided Approach for the Enhancement of Low-Lighting Underwater Images
  • Dataset introduction:

---

2.Object Detection/Classification

A. General

• MOUSS data
  • Data： MOUSS
  • Dataset introduction:
• MBARI dataset
  • Data： MBARI
  • Dataset introduction:
• HabCam dataset
  • Data： HabCam
  • Dataset introduction:
• OUC-vision dataset
  • Data： OUC-vision
  • Dataset introduction:
• MARIS project dataset
  • Data： MARIS project
  • Dataset introduction:
• NOAA dataset
  • Data： NOAA
  • Dataset introduction:
• Aqualoc dataset
  • Data： NOAA
  • Paper： AQUALOC: An Underwater Dataset for Visual-Inertial-Pressure Localization
  • Dataset introduction:
• Brackish dataset
  • Data： Brackish
  • Paper： Detection of Marine Animals in a New Underwater Dataset with Varying Visibility
  • Dataset introduction:
• SUN dataset
  • Data： SUN
  • Dataset introduction:

B. Human-robot cooperation

• Diver detection
  • Data： Diver detection
  • Paper： Toward a Generic Diver-Following Algorithm: Balancing Robustness and Efficiency in Deep Visual Detection
  • Dataset introduction:
• Robot tracking by detection
  • Data： Robot tracking by detection
  • Paper： Underwater multi-robot convoying using visual tracking by detection
  • Dataset introduction:
• CADDY diver pose data
  • Data： CADDY diver pose data
  • Paper： CADDY Underwater Stereo-Vision Dataset for Human–Robot Interaction (HRI) in the Context of Diver Activities
  • Dataset introduction:

C. Coral-reef

• UCSD
  • Data： UCSD
  • Paper： Automated annotation of coral reef survey images
  • Dataset introduction:
• Coral-reef Puerto Rico
  • Data： Coral-reef Puerto Rico
  • Dataset introduction:
• Coral-Net
  • Data： Coral-Net
  • Dataset introduction:

D. Fish

• WildFish database
  • Data： WildFish database
  • Paper： WildFish: A Large Benchmark for Fish Recognition in the Wild
  • Dataset introduction:
• Labeled fishes
  • Data： Labeled fishes
  • Paper： Automated Detection of Rockfish in Unconstrained Underwater Videos Using Haar Cascades and a New Image Dataset: Labeled Fishes in the Wild
  • Dataset introduction:
• Fish4Knowledge
  • Data： Fish4Knowledge
  • Dataset introduction:
• Fish database
  • Data： Fish database
  • Dataset introduction:
• AQUALIFEIMAGES database
  • Data： AQUALIFEIMAGES
  • Dataset introduction:
• Rockfish database
  • Data： Rockfish
  • Dataset introduction:
• Oceanwide database
  • Data： Oceanwide
  • Dataset introduction:
• Fish recognition
  • Data： Fish recognition
  • Paper： Supporting Ground-Truth annotation of image datasets using clustering
  • Dataset introduction:
• Fish detection and tracking
  • Data： Fish detection and tracking
  • Paper： An innovative web-based collaborative platform for video annotation
  • Dataset introduction:
• Fish trajectory detection
  • Data： Fish detection and tracking
  • Paper： Detection of Abnormal Fish Trajectories Using a Clustering Based Hierarchical Classifier
  • Dataset introduction:

E. Trash and marine debris

• TrashCan
  • Data： TrashCan
  • Paper： TrashCan: A Semantically-Segmented Dataset towards Visual Detection of Marine Debris
  • Dataset introduction:
• Trash-ICRA19
  • Data： Trash-ICRA19
  • Paper： Robotic Detection of Marine Litter Using Deep Visual Detection Models
  • Dataset introduction:
• Deep-sea debris database
  • Data： Deep-sea debris database
  • Dataset introduction:
• Tiny plastics posing threat to turtles
  • Data： Tiny plastics posing threat to turtles
  • Dataset introduction:

---

3. Acoustic Data

• Five-element acoustic dataset
  • Data： Five-element acoustic dataset
  • Paper： University of Texas Applied Research Laboratory Nov. 2009 Five-Element Acoustic Underwater Dataset
  • Dataset introduction:
• DIDSON dataset
  • Data： DIDSON dataset1 DIDSON dataset2 DIDSON dataset3
  • Paper： An underwater observation dataset for fish classification and fishery assessment
  • Dataset introduction:
• Spectrogram Analysis
  • Data： Spectrogram Analysis
  • Paper： On the detection of tracks in spectrogram images
  • Dataset introduction:
• Caves sonar and vision data
  • Data： Caves sonar and vision data
  • Paper： Underwater caves sonar data se
  • Dataset introduction:

---

4. Stereo Data

• Tasmania coral point
  • Data： Tasmania coral point
  • Paper： Towards autonomous habitat classification using Gaussian Mixture Models
  • Dataset introduction:
• Stereo from Flicker
  • Data： Stereo from Flicker
  • Paper： 3Deflicker from motion
  • Dataset introduction:
• CADDY stereo data
  • Data： CADDY stereo data
  • Paper： CADDY Underwater Stereo-Vision Dataset for Human–Robot Interaction (HRI) in the Context of Diver Activities
  • Dataset introduction:
• HIMB data for UWStereoNet
  • Data： HIMB data for UWStereoNet
  • Paper： UWStereoNet: Unsupervised Learning for Depth Estimation and Color Correction of Underwater Stereo Imagery
  • Dataset introduction:
• SQUID dataset
  • Data： SQUID dataset
  • Paper： Underwater Single Image Color Restoration Using Haze-Lines and a New Quantitative Dataset
  • Dataset introduction:

---

5. Image Segmentation

• SUIM
  • Data： SUIM
  • Paper： Semantic Segmentation of Underwater Imagery: Dataset and Benchmark
  • Code: Github
  • Dataset introduction:
• Coral-Net
  • Data： Coral-Net
  • Paper： CoralSeg: Learning coral segmentation from sparse annotations
  • Code: Github
  • Dataset introduction:
• Eilat dataset
  • Data： Eilat dataset
  • Paper： Improving Automated Annotation of Benthic Survey Images Using Wide-band Fluorescence
  • Dataset introduction:
• Change detection
  • Data： Change detection
  • Paper： Dataset on underwater change detection
  • Dataset introduction:

---

6.Underwater Datasets(include depth)

The underwater scene dataset in this part contains the depth map corresponding to the image

• Sea-thru dataset
  • Sea-thru dataset： Sea-thru
  • Paper： A Method For Removing Water From Underwater Images
  • Homepage：Homepage
  • Dataset introduction:

---

Contributing

If you are interested in contributing, please contact us by email.

---

Copyright notice

[Notes]This Github repo can be used by following the corresponding licenses. I want to emphasis that it may contain some PDFs or thesis, which were downloaded by me and can only be used for academic purposes. The copyrights of these materials are owned by corresponding publishers or organizations. All this are for better adademic research. If any of the authors or publishers have concerns, please contact me to delete or replace them.