Fast Underwater Image Enhancement for Improved Visual Perception

By M. Islam, Youya Xia, Junaed Sattar. IEEE Robotics and Automation Letters, 2019

1,454 citations recorded by the source corpus.

In this letter, we present a conditional generative adversarial network-based model for real-time underwater image enhancement. To supervise the adversarial training, we formulate an objective function that evaluates the perceptual image quality based on its global content, color, local texture, and style information. We also present EUVP, a large-scale dataset of a paired and an unpaired collection of underwater images (of ‘poor’ and ‘good’ quality) that are captured using seven different cameras over various visibility conditions during oceanic explorations and human-robot collaborative experiments. In addition, we perform several qualitative and quantitative evaluations which suggest that the proposed model can learn to enhance underwater image quality from both paired and unpaired training. More importantly, the enhanced images provide improved performances of standard models for underwater object detection, human pose estimation, and saliency prediction. These results validate that it is suitable for real-time preprocessing in the autonomy pipeline by visually-guided underwater robots. The model and associated training pipelines are available at https://github.com/xahidbuffon/funie-gan.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/ee56d8dc8d5900716d763d1cc94f3d226d8f9a25

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

1. Chongyi Li
2. Chunli Zhu
3. Liheng Bian
4. Lintao Peng
5. Peixian Zhuang
6. Chunle Guo
7. Runmin Cong
8. Sam Kwong

Citations in the corpus

• U-Shape Transformer for Underwater Image Enhancement (2021, methodology): “The enhancement result of FUnLE [19] is yellowish overall.”
• Using labeled data to evaluate change detectors in a multivariate streaming environment (2009, positive): “Most of them would adopt a backbone ResNet [7] , DeepLab [8] , Transformer [9] to extract meaningful features.”
• Simultaneous Enhancement and Super-Resolution of Underwater Imagery for Improved Visual Perception (2020, methodology): “Driven by large-scale supervised training [35, 69], these approaches learn sequences of non-linear filters to approximate the underlying pixel-to-pixel mapping [36] between the distorted and enhanced image domains.”
• Semantic Segmentation of Underwater Imagery: Dataset and Benchmark (2020, methodology): “First, the visual content of underwater imagery is entirely different because of the domainspecific object categories, background patterns, and optical distortion artifacts [10]; hence, the learning-based SOTA models trained on terrestrial data are not directly applicable.”
• Underwater Image Enhancement via Weighted Wavelet Visual Perception Fusion (2024, methodology): “In terms of contrast enhancement, GDCP [26], DTVR [6], GIFM [31], FUnIE-GAN [53], and SGUIE-Net [2] are weaker than UIECˆ2-Net [50] and PUIE-Net [55].”
• PUGAN: Physical Model-Guided Underwater Image Enhancement Using GAN With Dual-Discriminators (2023, methodology): “The UFO-120 dataset [26] contains 1500 annotated samples for large-scale training, and”