NanoDet Model‐Based Tracking and Inspection of Net Cage Using ROV

With the continuous and intensive development of ocean exploration, ROV (Remote Operated Vehicle) has become an important underwater carrier equipment and is widely used in underwater terrain survey and underwater target capture, etc. In this paper, we establish an environmental target recognition system based on forward looking sonar to ensure the safe navigation of ROV. The system consists of two components: hardware structure and software module. The hardware structure contains the equipments required by the system and the connections among them. The software module is used to process sonar image, including image preprocessing, target detection and positioning and navigation decision making. The pool test is carried out to validate the efficiency of the target detection and positioning algorithm. Meanwhile, we conduct a sea trial using ROV with the environmental target recognition system, results show that the ROV can take actions in a safe manner based on the navigation decisions.

In this paper we describe a successful model of underwater munitions characterization using a suite of integrated technologies. Adapted from mature and extensively utilized land-based munitions assessment techniques, a comprehensive suite of technologies are employed. Project phases and technologies include; diver reconnaissance utilizing former Navy EOD personnel, multi-beam sonar for bathymetry, side scan sonar for seafloor target detection and classification, sub-bottom profiling for sediment characterization, stationary scanning sonar to define high resolution bottom conditions, diver visual searches, full-coverage geophysical mapping via vessel-deployed magnetometer arrays, remote operated vehicle (ROV) deployments for visual and sonar searches and target interrogation, and ROVs equipped with an electromagnetic induction sensor array for target detection and interrogation. Additionally, three integrated underwater positioning systems were used to support accurate collection and integration of datasets within a comprehensive on-site GIS. Application of the methodology at two sites will be discussed. The first site in Seattle Washington comprised a highly cluttered environment around a working pier with water depths between 12 and 21 m. The second site comprises 1170 Ha at a former aerial bombing range at Choctawhatchee Bay near Pensacola, FL with water depths from 1 to 12 m with moderate currents.

Changes in sedimentation, sediment characteristics, and benthic macrofaunal assemblages around marine cage culture under seasonal monsoon scales in a shallow-water bay in Taiwan

Spatial and temporal variation of microbial populations and microbial metabolic potential in a tropical marine cage-culture sediment system

Improving the efficacy of the pilot remotely operated vehicle (ROV) interaction through extensive training is paramount in reducing the duration, and thus expense, of ROV deployments. To complete training without sacrificing operational windows, ROV simulators can be used. Since the ROV tether, which provides power and telemetry, will at times dominate the ROV motion, the tether must be accurately modeled over the full duration of a simulated ROV maneuver. One aspect of the tether dynamics that remains relatively untouched is the modeling of tether self-contact, contact with other tethers, or entanglement. The aim of this work is to present a computationally efficient and accurate method of detecting tether collisions. To this end, a combinatorial global optimization method is first used to determine the approximate separation distance minima locations. Then, a local optimization scheme is used to find the exact separation distances and the locations of the closest points. The first combinatorial stage increases the speed at which the minima can be found. The minimum separation distance information and its change with respect to time can then be used to continuously determine whether a collision has occurred. If a collision is detected, a contact force is calculated from the interference geometry and applied at the collision site.

The scale and intensity of marine cage culture have increased in the Asian-Pacific region, particularly in oligotrophic waters where coral reef organisms flourish. In this study, the influence of marine cage culture on subtropical coral communities in turbid waters was evaluated by measuring environmental parameters and benthic community compositions at Magongwan in the Penghu Islands, Taiwan. A canonical discriminant analysis of environmental parameters revealed that elevated levels of ammonium, nitrite, and chlorophyll a (chl a) released from the cages were the main pollution indicators that, in addition to sedimentation and turbidity, distinguished Impact Zone 1 (cage-culture zone) from the other 2 zones - Impact Zone 2 (800 m away from the cages) and the reference zone - in these turbid waters. Results of the canonical correlation analysis indicated that the coverage extents of macroalgae, sponges, and zoanthids were strongly correlated with levels of ammonium, nitrite, phosphate, chl a, and dissolved oxygen. Coral communities in Impact Zone 1 were mostly composed of stress-tolerant massive and submassive corals, but were lacking branching Acropora corals. In contrast, coral communities in the other zones, with high habitat complexity and species richness, were dominated by coral species with diverse morphologies, including branching Acropora coral communities. These results suggest that marine cage culture has been causing chronic nutrient enrichment in the surrounding waters at Magongwan, which may have resulted in a deterioration of suitable habitats for coral reef organisms. Nevertheless, intermediate levels of nutrients and particulate organic matter relative to the other zones might have been caused by the adjacent cage culture, resulting in the high coral coverage and diversity in the Impact Zone 2 in particular.

Localization is an important and extensively studied problem in ad-hoc wireless sensor networks. Given the connectivity graph of the sensor nodes,along with additional local information (e.g. distances, angles, orientations etc.), the goal is to reconstruct the global geometry of the network. In this paper, we study the problem of localization with noisy distance and angle information. With no noise at all, the localization problem with both angle (with orientation) and distance information is trivial. However, in the presence of even a small amount of noise, we prove that the localization problem is NP hard.Localization with accurate distance information and relative angle information is also hard. These hardness results motivate our study of approximation schemes. We relax the non-convex constraints to approximating convex constraints and propose linear programs (LP) for two formulations of the resulting localization problem, which we call the weak deployment and strong deployment problems.These two formulations give upper and lower bounds on the location uncertainty respectively: No sensor is located outside its weak deployment region, and each sensor can be anywhere in its strong deployment region without violating the approximate distance and angle constraints. Though LP-based algorithms are usually solved by centralized methods, we propose distributed, iterative methods, which are provably convergent to the centralized algorithm solutions. We give simulation results for the distributed algorithms, evaluating the convergence rate, dependence on measurement noises,and robustness to link dynamics.

Object detection/tracking toward underwater photographs by remotely operated vehicles (ROVs)

Adaptive PD-controller for positioning of a remotely operated vehicle close to an underwater structure: Theory and experiments

This paper proposes a vector-based localization system that uses both distance and angle information. In wireless sensor networks, the positions of nodes are commonly determined by a range-based localization system using distance information. If both distance and angle information are available, it is possible to improve the accuracy of estimating the positions of nodes compared to a positioning system with only distance information. Existing studies using distance and angle information assume that all the nodes are directly connected to one another and do not consider a method for measuring angle information between the nodes that are not directly connected. However, this assumption may not be valid for realworld wireless sensor networks especially with a large number of nodes having a limited communication range. The proposed localization algorithm solves this problem by a vector combination that transforms the vectors on the local coordinate system to the network-wide global coordinate system. The proposed algorithm is shown to be robust especially even in a network with 1-edge connectivity. Simulation results show that the proposed algorithm has up to 70% higher positioning accuracy compared to the existing iterative range-based algorithm such as MDS-MAP(C,R).

Teleoperating Remotely Operated Vehicles (ROV) underwater is a challenging and tedious task as the operators are physically removed from the sites and their situational awareness is low due to restricted camera views, poor visibility under water, and limited three dimensional perception. This paper describes three vision system technologies, developed originally for space applications, which will enable more efficient ROV operations. Automatic detection and tracking of artificial and natural targets will enable autonomous ROV station-keeping (mooring) allowing the operators to focus on their main tasks. The 3D modeling system will create instantly models of underwater worksites and seafloor increasing thus operators' spatial perception and allowing measurements and precise manipulation tasks. Initial results of underwater experiments are presented.

Various underwater munitions response applications, including UXO site investigation and remediation, and mine countermeasures and infrastructure characterization, require the detection, geo-registration, and characterization of targets on, or below, the seafloor. The need to find explosive items of interest or high value infrastructure obscured in coastal and marine sediments has motivated the need for marine controlled source electromagnetic technologies integrated with remotely operated platforms. We have developed and demonstrated multiple 3D time-domain EM systems from various platforms to optimize the utility of underwater digital geophysical mapping (DGM) and advanced geophysical classification (AGC). Often, the 3D electromagnetic sensing technology requires close standoff to the seafloor, which is enabled by specialized autonomous platforms and/or tightly controlled tow winching systems. When tightly controlled and integrated with advanced underwater positioning systems, the sensor array systems exhibit new capabilities for characterizing the location, size, shape, and extent of the compact metallic objects. This is an especially powerful tool for critical nearshore applications where acoustic and optical methods are significantly challenged. We present multiple implementations of 3DEM sensors from remotely operated vehicles (ROVs), bottom crawlers, and surface towed arrays. We present the results of testing on both small inspection-class ROVs in littoral waters and amphibious bottom-crawlers in challenging surf zone conditions as well as large working-class ROVs in deeper water. Presented concepts include full axial illumination using 3D transmitter configurations with a focus on the engineering tradeoffs between practical hydrodynamic towed implementations and full angular illumination for high quality underwater UXO classification.

The accurate target detection under different environmental conditions and the real-time target positioning are vital for the successful accomplishment of underwater missions of Remotely operated vehicles (ROVs). In this paper, we propose a vision-based underwater target detection and positioning approach to detect and estimate the position and attitude of artificial underwater targets. The proposed approach is composed of an underwater target detection algorithm YOLO-T and a target positioning algorithm. Firstly, we modify the structure of YOLOv5 algorithm using Ghost module and SE attention module to improve the calculation time of target detection. Secondly, a series of image processing operations are performed on the improved YOLOv5 detection results to increase the detection accuracy. Thirdly, a cooperative marker is designed as the artificial underwater target, and the corresponding positioning algorithm is presented to calculated the position and attitude of the target according to the geometric information of the designed marker. We validate our approach through experimental tests respectively in a water tank, an anechoic tank, and the sea trial in Huanghai Sea in China. The results demonstrate the accurate performance of the proposed detection and positioning method.

This paper presents a simple analog neuromorphic system in microelectronics suitable for an audio source localization problem. The receptor can find the relative position of an audio source by the information of angles and distances according to acquire acoustic signals. In this paper, we focus on the angle detection, but some information of the distance is also presented. This paper also presents the development of some alternatives to the most important circuit blocks in neuromorphic systems: the neurons and the synapses. The results are validated in two different levels: the system level and the transistor level.

Remotely-Operated Vehicles (ROVs) serve many functions during site characterization surveys including dive team assistance, quality control, and anomaly reacquisition. Reacquisition applications require detection, geo-registration, and investigation of items of interest located on, or below, the seafloor. Underwater target detection using ROVs poses many challenges. Detection sensors must be mounted far enough away from the vehicle to reduce noise effects from thrusters and other vehicle sensors. Conversely, detection sensors must be mounted close enough to the vehicle to maintain vehicle maneuverability. Vehicle noise characterization with respect to sensor standoff and orientation is discussed. ROV applications require the collection of correlated detection sensor data along with global position information. Positional accuracy varies depending on the method used and can be dependent on site and deployment conditions. Positioning systems, including Ultra-Short Base Line (USBL), Doppler Velocity Logs (DVL), and Real-time Kinematic Global Positioning System (RTK-GPS) are discussed with respect to underwater object localization. We present the development of underwater metal detection systems consisting of magnetometer and electromagnetic induction (EMI) sensors mounted on a mini-ROV. Detection sensors include fluxgate magnetometers and a diver handheld EMI sensor modified for data transmission through the ROV umbilical for digital data capture and audio data delivery topside. Sensor sensitivities are defined through analysis of data collected against targets of various sizes. We discuss the development, performance, and operation of various payloads for different applications and compare ROV and dive team results and efficiency.