25 January 2022 in IEEE Sensors Journal
Robust underwater imaging sensors are essential in applications such as identifying and classifying marine animals in turbid waters, typically found around coastal monitoring sites. The Unobtrusive Multi-static Serial LiDAR Imager (UMSLI) has been developed to capture images taken in degraded underwater environments and provides superior range and image contrast over conventional optical cameras. In contrast to traditional underwater LiDAR systems that employ blue-green lasers, UMLSI will not harm the vision of marine animals, which is an important aspect in such settings. As with any other underwater LiDAR sensors, improving the quality of images taken in highly turbid water is essential to the UMSLI system. This paper proposes a novel machine learning image enhancement technique based on a Generative Adversarial Network (GAN) framework. One main contribution of the method is the incorporation of a correntropy-based perceptual loss. This technique has shown to be effective in enhancing ground-based images and has been adapted for LiDAR image enhancement. Given the limitation of existing underwater LiDAR data, a method for simulating degraded data for training is also presented. This LiDAR technique was validated using LiDAR data captured by the UMSLI system within Florida Atlantic University’s Harbor Branch Oceanographic Institute (FAU-HBOI) optical test facility and at the Marine and Coastal Research Laboratory at the Pacific Northwest National Laboratory (MCRL-PNNL). This image enhancement technique can be readily extended to other underwater LiDAR systems.
1 January 2023 in Journal of Applied Remote Sensing
Imaging through degraded visual environments is a challenging task in remote sensing missions. Image degradation may come from the loss of contrast due to particle scattering and/or distortion due to turbulence-induced effects. The problem is especially challenging when imaging from moving platforms such as autonomous underwater vehicles. One potential approach to address these issues is to use multiple images and employ a multi-frame image fusion technique to aid in the recovery of corrupted image detail and quality. A machine learning (ML)-based image enhancement and fusion technique is investigated to restore images distorted by underwater turbulence. The main contributions include the incorporation of a ML-based image weight predictor to predict the ideal weight maps to be used in an image fusion process. This network is trained using a generative adversarial network framework and a synthetically generated image dataset, which is created according to an analytical image degradation model. In addition, the image loss function for the weight map predictor is determined by the final fused image, resulting in a balanced fusion technique that can reduce image distortion, recover crisp image detail, and reduce the overall noise figure. Another key contribution of this paper is the adoption of an image loss function that incorporates an innovative combined correntropy and Fourier space loss function to reinforce the network in both the spatial and frequency domain. The performance of the proposed algorithm is evaluated using a synthetic validation dataset of images and several real datasets captured at the Naval Research Lab’s Simulated Turbulence and Turbidity Environment under various controlled turbulence intensities.