Search Results (23)

Coral reefs are among the most important and vulnerable marine ecosystems worldwide. AI-powered underwater visual monitoring has become essential for effective reef conservation, yet current methods still face severe limitations: spectral ambiguity caused by underwater turbidity, fine-grained confusion in early coral health assessment, and discrete forecasting models that cannot represent continuous ecological degradation dynamics. To address these issues, we propose Spectral-YOLOv13, a dual-domain vision-Mamba sensing framework for high-precision coral health evaluation and continuous ecological forecasting. The framework incorporates three novel components: a Wavelet-Integrated Omni-Neck (WIO-Neck) to perform multi-scale spectral filtering and suppress turbidity-induced noise; a Contrastive Prototype Head (CP-Head) to enhance discriminability between visually similar health states; and a Bio-Mamba Predictor based on state-space models to capture long-term continuous health trajectories. Extensive experiments on the CR-Mix++ dataset demonstrate that Spectral-YOLOv13 achieves 53.8% mAP with strong robustness in turbid underwater environments. It reduces four-week forecasting error by 26.8% and maintains real-time inference speed at 112 FPS. This work provides a reliable and high-performance vision framework for practical underwater coral reef monitoring and proactive conservation management. Full article

Bait costs constitute 40–50% of the total expenditure in river crab aquaculture, highlighting the critical need for accurately assessing crab growth and scientifically determining optimal feeding regimes across different farming stages. Current traditional methods rely on periodic manual sampling to monitor growth status and artificial feeding platforms to observe consumption and adjust bait input. These approaches are inefficient, disruptive to crab growth, and fail to provide comprehensive growth data. Therefore, this study proposes a machine vision-based monitoring system for river crab feeding platforms. Firstly, the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm is applied to enhance underwater images of river crabs. Subsequently, an improved YOLOv11 (You Only Look Once) model is introduced and applied for multi-target detection and counting in crab ponds, enabling the extraction of information related to both river crabs and bait. Concurrently, underwater environmental parameters are monitored in real-time via an integrated environmental information sensing system. Finally, an information processing platform is established to facilitate data sharing under a “detection–processing–distribution” workflow. The real crab farm experimental results show that the river crab quality error rate was below 9.57%, while the detection rates for both corn and pellet baits consistently exceeded 90% across varying conditions. These results indicate that the proposed system significantly enhances farming efficiency, elevates the level of automation, and provides technological support for the river crab aquaculture industry. Full article

Light, as one of the most crucial environmental factors, plays an essential role in the growth, physiology, and evolutionary survival of fish. To cope with diverse light conditions in aquatic environments, fish adapt through photosensory systems composed of both visual and non-visual pathways. The retina is a key component of the visual system of fish, capable of converting external optical signals into neural electrical signals, making it crucial for visual formation. During the process of visual signal transduction, opsins serve as the molecular foundation for vision formation. They can be divided into two major categories: visual opsins and non-visual opsins. Among these, melanopsin, as a member of the non-visual opsin family, acts as a key upstream factor in the circadian phototransduction pathway of fish. In this review, we review the adaptability of fish retinal structures to light reception and introduce in detail the gene diversity and relative expression levels of fish opsins. At the same time, we comprehensively describe the molecular mechanism by which fish adapt to changes in the underwater light environment. We also concluded that melanopsin, as a non-imaging photoreceptor, possesses not only core light-sensing functions but also non-imaging visual functions such as circadian rhythm regulation, body coloration changes, and hormone secretion. This review suggests that future research should not only elucidate the physiological functions of melanopsin in fish but also comprehensively reveal the mechanisms underlying the multi-adaptive nature of fish vision across varying light environments. Through these studies, researchers can have a deeper understanding of the physiological regulation mechanism of fish in complex light environments, and then formulate fish light environment management strategies, optimize aquaculture practices, improve economic returns, and promote the development of related fields. Full article

Autonomous underwater cleaning in water pools requires reliable perception, efficient coverage path planning, and robust control. However, existing autonomous underwater vehicle (AUV) cleaning systems often suffer from fragmented software frameworks that limit end-to-end performance. To address these challenges, this paper proposes an integrated vision-based autonomous underwater cleaning system that combines global-camera AprilTag localization, YOLOv8-based dirt detection, and a multi-scale A* coverage path planning algorithm. The perception and planning modules run on a host computer system, while a NanoPi-based controller executes motion commands through a lightweight JSON-RPC protocol over Ethernet. This architecture ensures real-time coordination between visual sensing, planning, and hierarchical control. Experiments conducted in a simulated pool environment demonstrate that the proposed system achieves accurate localization, efficient planning, and reliable cleaning without blind spots. The results highlight the effectiveness of integrating vision, multi-scale planning, and lightweight embedded control for autonomous underwater cleaning tasks. Full article

The integrity of subsea oil and gas pipelines is essential for offshore safety and environmental protection. Conventional leak detection approaches, such as manual inspection and indirect sensing, are often costly, time-consuming, and prone to subjectivity, motivating the development of automated methods. In this study, we present a deep learning-based framework for detecting underwater leaks using images acquired in controlled experiments designed to reproduce representative conditions of subsea monitoring. The dataset was generated by simulating both gas and liquid leaks in a water tank environment, under scenarios that mimic challenges observed during Remotely Operated Vehicle (ROV) inspections along the Brazilian coast. It was further complemented with artificially generated synthetic images (Stable Diffusion) and publicly available subsea imagery. Multiple Convolutional Neural Network (CNN) architectures, including VGG16, ResNet50, InceptionV3, DenseNet121, InceptionResNetV2, EfficientNetB0, and a lightweight custom CNN, were trained with transfer learning and evaluated on validation and blind test sets. The best-performing models achieved stable performance during training and validation, with macro F1-scores above 0.80, and demonstrated improved generalization compared to traditional baselines such as VGG16. In blind testing, InceptionV3 achieved the most balanced performance across the three classes when trained with synthetic data and augmentation. The study demonstrates the feasibility of applying CNNs for vision-based leak detection in complex underwater environments. A key contribution is the release of a novel experimentally generated dataset, which supports reproducibility and establishes a benchmark for advancing automated subsea inspection methods. Full article

Research in marine hydrographic environmental monitoring continues to deepen, necessitating a hardware platform capable of traversing air–water interfaces to collect vertical gradient parameters across oceanographic profiles. This paper proposes a deeply integrated heterogeneous monitoring platform for marine hydrological vertical profiling, addressing the functional limitations of conventional unmanned surface vehicles (USVs) and unmanned aerial vehicles (UAVs) in subsurface monitoring. By co-designing a hybrid aerial underwater vehicle (HAUV) with cross-domain capabilities and a USV, the system leverages USVs for long-endurance surface operations and HAUVs for high-speed vertical column monitoring. Key innovations include (1) a distributed collaborative architecture enabling “Air–Sea–Air” cyclic operations; (2) dynamic modeling of HAUV-USV interactions incorporating aerodynamic and hydrodynamic coupling; (3) an MPC-based collaborative tracking algorithm for real-time USV pursuit under marine disturbances; and (4) a vision-guided synchronous landing strategy achieving decimeter-level docking accuracy in bad conditions. Simulation experiments validate the system’s efficacy in trajectory tracking and precision landing. This work bridges the critical gap in marine vertical profile monitoring while demonstrating robust cross-domain coordination. Full article

Underwater navigation remains constrained by technological limitations, driving the exploration of alternative approaches such as polarized light-based systems. This review systematically examines advances in polarized navigation from three perspectives. First, the principles of atmospheric polarization navigation are analyzed, with their operational mechanisms, advantages, and inherent constraints dissected. Second, innovations in bionic polarization multi-sensor fusion positioning are consolidated, highlighting progress beyond conventional heading-direction extraction. Third, emerging underwater polarization navigation techniques are critically evaluated, revealing that current methods predominantly adapt atmospheric frameworks enhanced by advanced filtering to mitigate underwater interference. A comprehensive synthesis of underwater polarization modeling methodologies is provided, categorizing physical, data-driven, and hybrid approaches. Through rigorous analysis of studies, three persistent barriers are identified: (1) inadequate polarization pattern modeling under dynamic cross-media conditions; (2) insufficient robustness against turbidity-induced noise; (3) immature integration of polarization vision with sonar/IMU (Inertial Measurement Unit) sensing. Targeted research directions are proposed, including adaptive deep learning models, multi-spectral polarization sensing, and bio-inspired sensor fusion architectures. These insights establish a roadmap for developing reliable underwater navigation systems that transcend current technological boundaries. Full article

Harbor seals (Phoca vitulina) have excellent perception of water disturbances and can still sense targets as far as 180 m away, even when they lose their vision and hearing. This exceptional capability is attributed to the undulating structure of its vibrissae. These specialized whiskers not only effectively suppress vortex-induced vibrations (VIVs) during locomotion but also amplify the vortex street signals generated by the wake of a target, thereby enhancing the signal-to-noise ratio (SNR). In recent years, researchers in fluid mechanics, bionics, and sensory biology have focused on analyzing the hydrodynamic characteristics of seal vibrissae. Based on bionic principles, various underwater biomimetic seal whisker sensors have been developed that mimic this unique geometry. This review comprehensively discusses research on the hydrodynamic properties of seal whiskers, the construction of three-dimensional geometric models, the theoretical foundations of fluid–structure interactions, the advantages and engineering applications of seal whisker structures in suppressing VIVs, and the design of sensors inspired by bionic principles. Full article

Camera-centric perception has matured into a cornerstone of modern autonomy, from self-driving cars and factory cobots to underwater and planetary exploration. This review synthesizes more than a decade of progress in vision-based robotic navigation through an engineering lens, charting the full pipeline from sensing to deployment. We first examine the expanding sensor palette—monocular and multi-camera rigs, stereo and RGB-D devices, LiDAR–camera hybrids, event cameras, and infrared systems—highlighting the complementary operating envelopes and the rise of learning-based depth inference. The advances in visual localization and mapping are then analyzed, contrasting sparse and dense SLAM approaches, as well as monocular, stereo, and visual–inertial formulations. Additional topics include loop closure, semantic mapping, and LiDAR–visual–inertial fusion, which enables drift-free operation in dynamic environments. Building on these foundations, we review the navigation and control strategies, spanning classical planning, reinforcement and imitation learning, hybrid topological–metric memories, and emerging visual language guidance. Application case studies—autonomous driving, industrial manipulation, autonomous underwater vehicles, planetary rovers, aerial drones, and humanoids—demonstrate how tailored sensor suites and algorithms meet domain-specific constraints. Finally, the future research trajectories are distilled: generative AI for synthetic training data and scene completion; high-density 3D perception with solid-state LiDAR and neural implicit representations; event-based vision for ultra-fast control; and human-centric autonomy in next-generation robots. By providing a unified taxonomy, a comparative analysis, and engineering guidelines, this review aims to inform researchers and practitioners designing robust, scalable, vision-driven robotic systems. Full article

Maritime vision sensing can be applied to maritime unmanned systems to perform search and rescue (SAR) missions under complex marine environments, as multiple unmanned aerial vehicles (UAVs) and unmanned surface vehicles (USVs) are able to conduct vision sensing through the air, the water-surface, and underwater. However, in these vision-based maritime SAR systems, collaboration between UAVs and USVs is a critical issue for successful SAR operations. To address this challenge, in this paper, we propose a long-endurance collaborative SAR scheme which exploits the complementary strengths of the maritime unmanned systems. In this scheme, a swarm of UAVs leverages a multi-agent reinforcement-learning (MARL) method and probability maps to perform cooperative first-phase search exploiting UAV’s high altitude and wide field of view of vision sensing. Then, multiple USVs conduct precise real-time second-phase operations by refining the probabilistic map. To deal with the energy constraints of UAVs and perform long-endurance collaborative SAR missions, a multi-USV charging scheduling method is proposed based on MARL to prolong the UAVs’ flight time. Through extensive simulations, the experimental results verified the effectiveness of the proposed scheme and long-endurance search capabilities. Full article

Acoustic and optical sensing modalities represent two of the primary sensing methods within underwater environments, and both have been researched extensively in previous works. Acoustic sensing is the premier method due to its high transmissivity in water and its relative immunity to environmental factors such as water clarity. Optical sensing is, however, valuable for many operational and inspection tasks and is readily understood by human operators. In this work, we quantify and compare the operational characteristics and environmental effects of turbidity and illumination on two commercial-off-the-shelf sensors and an additional augmented optical method, including: a high-frequency, forward-looking inspection sonar, a stereo camera with built-in stereo depth estimation, and color imaging, where a laser has been added for distance triangulation. The sensors have been compared in a controlled underwater environment with known target objects to ascertain quantitative operation performance, and it is shown that optical stereo depth estimation and laser triangulation operate satisfactorily at low and medium turbidites up to a distance of approximately one meter, with an error below 2 cm and 12 cm, respectively; acoustic measurements are almost completely unaffected up to two meters under high turbidity, with an error below 5 cm. Moreover, the stereo vision algorithm is slightly more robust than laser-line triangulation across turbidity and lighting conditions. Future work will concern the improvement of the stereo reconstruction and laser triangulation by algorithm enhancement and the fusion of the two sensing modalities. Full article

Underwater imaging has been present for many decades due to its relevance in vision and navigation systems. In recent years, advances in robotics have led to the availability of autonomous or unmanned underwater vehicles (AUVs, UUVs). Despite the rapid development of new studies and promising algorithms in this field, there is currently a lack of research toward standardized, general-approach proposals. This issue has been stated in the literature as a limiting factor to be addressed in the future. The key starting point of this work is to identify a synergistic effect between professional photography and scientific fields by analyzing image acquisition issues. Subsequently, we discuss underwater image enhancement and quality assessment, image mosaicking and algorithmic concerns as the last processing step. In this line, statistics about 120 AUV articles fro recent decades have been analyzed, with a special focus on state-of-the-art papers from recent years. Therefore, the aim of this paper is to identify critical issues in autonomous underwater vehicles encompassing the entire process, starting from optical issues in image sensing and ending with some issues related to algorithmic processing. In addition, a global underwater workflow is proposed, extracting future requirements, outcome effects and new perspectives in this context. Full article

Nowadays, more frequently, it is necessary to perform underwater operations such as surveying an area or inspecting and intervening on industrial infrastructures such as offshore oil and gas rigs or pipeline networks. Recently, the use of Autonomous Underwater Vehicles (AUV) has grown as a way to automate these tasks, reducing risks and execution time. One of the used sensing modalities is vision, providing RGB high-quality information in the mid to low range, making it appropriate for manipulation or detail inspection tasks. This work presents the use of a deep neural network to perform pixel-wise 3D segmentation of pipes and valves on underwater point clouds generated using a stereo pair of cameras. In addition, two novel algorithms are built to extract information from the detected instances, providing pipe vectors, gripping points, the position of structural elements such as elbows or connections, and valve type and orientation. The information extracted on spatially referenced point clouds can be unified to form an information map of an inspected area. Results show outstanding performance on the network segmentation task, achieving a mean F1-score value of 88.0% at a pixel-wise level and of 95.3% at an instance level. The information extraction algorithm also showcased excellent metrics when extracting information from pipe instances and their structural elements and good enough metrics when extracting data from valves. Finally, the neural network and information algorithms are implemented on an AUV and executed in real-time, validating that the output information stream frame rate of 0.72 fps is high enough to perform manipulation tasks and to ensure full seabed coverage during inspection tasks. The used dataset, along with a trained model and the information algorithms, are provided to the scientific community. Full article

Aquaculture net pens inspection and monitoring are important to ensure net stability and fish health in the fish farms. Remotely operated vehicles (ROVs) offer a low-cost and sophisticated solution for the regular inspection of the underwater fish net pens due to their ability of visual sensing and autonomy in a challenging and dynamic aquaculture environment. In this paper, we report the integration of an ROV with a visual servoing scheme for regular inspection and tracking of the net pens. We propose a vision-based positioning scheme that consists of an object detector, a pose generator, and a closed-loop controller. The system employs a modular approach that first utilizes two easily identifiable parallel ropes attached to the net for image processing through traditional computer vision methods. Second, the reference positions of the ROV relative to the net plane are extracted on the basis of a vision triangulation method. Third, a closed-loop control law is employed to instruct the vehicle to traverse from top to bottom along the net plane to inspect its status. The proposed vision-based scheme has been implemented and tested both through simulations and field experiments. The extensive experimental results have allowed the assessment of the performance of the scheme that resulted satisfactorily and can supplement the traditional aquaculture net pens inspection and tracking systems. Full article

In an endeavor to study natural systems at multiple spatial and taxonomic resolutions, there is an urgent need for automated, high-throughput frameworks that can handle plethora of information. The coalescence of remote-sensing, computer-vision, and deep-learning elicits a new era in ecological research. However, in complex systems, such as marine-benthic habitats, key ecological processes still remain enigmatic due to the lack of cross-scale automated approaches (mms to kms) for community structure analysis. We address this gap by working towards scalable and comprehensive photogrammetric surveys, tackling the profound challenges of full semantic segmentation and 3D grid definition. Full semantic segmentation (where every pixel is classified) is extremely labour-intensive and difficult to achieve using manual labeling. We propose using label-augmentation, i.e., propagation of sparse manual labels, to accelerate the task of full segmentation of photomosaics. Photomosaics are synthetic images generated from a projected point-of-view of a 3D model. In the lack of navigation sensors (e.g., a diver-held camera), it is difficult to repeatably determine the slope-angle of a 3D map. We show this is especially important in complex topographical settings, prevalent in coral-reefs. Specifically, we evaluate our approach on benthic habitats, in three different environments in the challenging underwater domain. Our approach for label-augmentation shows human-level accuracy in full segmentation of photomosaics using labeling as sparse as $0.1\%$, evaluated on several ecological measures. Moreover, we found that grid definition using a leveler improves the consistency in community-metrics obtained due to occlusions and topology (angle and distance between objects), and that we were able to standardise the 3D transformation with two percent error in size measurements. By significantly easing the annotation process for full segmentation and standardizing the 3D grid definition we present a semantic mapping methodology enabling change-detection, which is practical, swift, and cost-effective. Our workflow enables repeatable surveys without permanent markers and specialized mapping gear, useful for research and monitoring, and our code is available online. Additionally, we release the Benthos data-set, fully manually labeled photomosaics from three oceanic environments with over 4500 segmented objects useful for research in computer-vision and marine ecology. Full article