Satellite imagery-based ship detection is indispensable in maritime surveillance and monitoring the naval activities. Machine learning is an effective approach that enables the process to be automatic and more accurate as compared to many other approaches. Generally, optical and synthetic aperture radar satellite images are often employed for detecting/locating various marine activities using different methods. However, models trained on one set of images often yield large uncertainties when testing on other sets of images due to the complex scene characteristics. This study proposes a novel lightweight computationally efficient deep learning-based general ship detection model called the Multi- Attentive General Ship Detector (MAGSD) for detecting ships in both optical and SAR satellite images. The model is trained with the SAR Ship Dataset (SDD), which has ship instances from Gaofen-3 and Sentinel-1 SAR satellite images, and the MASATI dataset that contains ship instances from the Microsoft Bing Map. The proposed model focuses on the attention-guided convolutional neural network for extracting feature maps for detection, which bridges the gap between SAR and optical image characteristics constraints by focusing on different levels of convolutional features in the network. The model is built with a novel feature extractor that has fourteen convolutional layers with six max pool layers and six attention layers, connecting several convolutional points to focus on local features in different depth maps which serve as the backbone of the model. The comparative analysis showed the robustness of the proposed model over the state-of-the-art baseline model YOLOv5s, with a precision of 8.2% and a recall of 9.63%. These results indicate that the proposed model holds the potential to serve as an efficient tool for ship detection in any satellite images and contributes to the enhanced coastal surveillance and bolsters global naval security.
Underwater radiometers are generally calibrated in air using a standard source. The immersion factors are required for these radiometers to account for the change in the in-water measurements with respect to in-air due to the different refractive index of the medium. The immersion factors previously determined for RAMSES series of commercial radiometers manufactured by TriOS are applicable to clear oceanic waters. In typical inland and turbid productive coastal waters, these experimentally determined immersion factors yield significantly large errors in water-leaving radiances (Lw) and hence remote sensing reflectances (Rrs). To overcome this limitation, a semi-analytical method with based on the refractive index approximation is proposed in this study, with the aim of obtaining reliable Lw and Rrs from RAMSES radiometers for turbid and productive waters within coastal and inland water environments. We also briefly show the variation of pure water immersion factors (Ifw) and newly derived If on Lw and Rrs for clear and turbid waters. The remnant problems other than the immersion factor coefficients such as transmission, air-water and water-air Fresnel’s reflectances are also discussed.
The assessment of water clarity of any regional water body is particularly important from ecological and water quality perspectives, especially in the regions which are highly influenced by sediment run-off and seasonal fluctuations in turbidity. The ocean colour remote sensing has played a significant role in monitoring the turbidity level in marine and inland water bodies. However, algorithms to accurately estimate the turbidity in such optically complex waters are scarce or limited by high level of uncertainty due to various issues. The present study proposes a simple, two band algorithm to estimate turbidity in both turbid and clear waters. It was found that the band ratio of remote sensing reflectance (Rrs(670)/Rrs(670)+Rrs(555)) represents the proxy of TSS (Total suspended sediment) and therefore, positively correlates to turbidity. The new algorithm is based on the assumption that light reflected in these two vital bands contains the essential information regarding the total suspended matter in the water column. The statistical results showed that the percent mean relative error between the predicted turbidity and the measured turbidity was within ±20%. To further demonstrate the robustness of the present algorithm, the spatial grid contours for the measured and the predicted turbidity was generated for the month of January 2014, August 2013 and May 2012 for the coastal waters in Bay of Bengal (Point Calimere, located in the southeast coast of India). The close consistency between the predicted and measured turbidity spatial patterns revealed that the present algorithm can be applied with high confidence to predict turbidity in both coastal and inland waters.
Scattering phase function plays a crucial role in studies and calculations based on radiative transfer theory in water as well as atmosphere. A model based on Mie theory is developed for estimating the particulates-in-water scattering phase function for forward angles (0.1° - 90°). Particle size distribution (PSD) slope (ξ) and bulk refractive index (n) are chosen as key inputs for this proposed model. The PSD slope can be estimated from the attenuation spectrum measured directly in-situ and the bulk refractive index can be calculated by an inversion model using measured backscattering ratio (BP) and PSD slope. The attenuation spectrum and backscattering ratio can be easily measured in-situ using commercially available instruments in real time. The entire range of forward angles is divided into two ranges and phase function is modeled separately in the ranges 0.1° - 5° and 5° - 90°, from numerically calculated Volume Scattering Function (VSF) using Mie theory. The division boundary is decided owing to the fact that the scattering phase functions, for different oceanic conditions, exhibit a change in slope at approximately 5°. Performance of the present model is evaluated by comparing with existing empirical and analytical models as well as measured phase functions. The proposed phase function model shows a considerable improvement upon existing models, and will have important applications in remote sensing applications and underwater studies.
Ocean Color Monitor-2 (OCM-2) on-board Oceansat 2 satellite is a multi-spectral sensor with a spatial resolution of 360×250m. Despite the presence of improved spatial resolution for better ocean color interpretation within coastal zones; differences among the OCM-2 detectors lead to striping artifacts in the along-track direction limiting the ocean color observations. Existing calibration methods do not characterize the striping noise efficiently. Destriping algorithms are generally applied to Level 2 radiance or biogeochemical products (i.e., post-radiometric and atmospheric correction), to remove the striping artifacts in order to ensure quality products. The present study focuses to reveal a robust method which effectively removes the striping effects in the TOA radiance products. Preliminary results obtained from this approach have been highlighted which show significant improvement in image quality for Level 1B (TOA radiance) and Level 2 (Water leaving radiance (Lw) and biogeochemical) products. The proposed method operates on a pixel by pixel basis with an aim to maintain the spatial and spectral resolution of data and ensure image quality in the derived products.
SeaWiFS RCA-Chl along with sea surface height variations/geostrophic currents, sea surface temperature, wind
speed/direction and field observation data, are used to first describe comprehensively the occurrences of various
hazardous algal blooms (HABs) and their underlying mechanisms and link to nutrient enrichment during the summer in
shelf-slope waters off the Northwest Pacific (NWP). These datasets provide a coherent view of the summertime
evolution of HABs and related physical processes in four common dynamic regions: coastal cold/estuary water zones,
upwelling zones next to the coast, repeated meanders/eddies, and frontal regimes induced by the Kuroshio and its
tributaries. High blooms coincided with the coastal upwelling and cyclonic eddy regimes that followed SST minimum
and large negative SSH along with favorable phase of winds. By contrast, relatively low mean RCA were consistent with
the fronts and anticyclonic meanders revealing moderate-high SSH fields along with variable winds blown off the NWP
coast. These anticyclonic meanders, on some occasions, when nutrient-containing coastal water setoff higher chlorophyll
biomass and major currents gained force in August, straddled the continental margin, entraining high chlorophyll water
from the coast and from the adjacent cyclonic eddies located nearby into their outer rings that formed a conveyer-belt
system of transport to inject coastal blooms into the deep-sea (e.g., East Sea) region of the NWP. The above findings
based on satellite data combined with field hydrographic/ bloom observation data evidently illustrated richness of the
response of summer HABs to the surface circulation and nutrient enrichment processes in shelf-slope waters off the
NWP coast.
The spatial and temporal distributional patterns of suspended sediments (SS) in the East China Sea (ECS) and Yellow
Sea (YS) were investigated by using satellite ocean color data from SeaWiFS and by using in-situ data. Except for the
Southeastern YS, the overall distribution patterns of SS revealed a general, cross-shelf decreasing trend along the
sediment dispersal system away from the rivers, closely consistent with the previous classification of SS - Infant stage,
Younger stage, Mature stage and Old stage. We hypothesize that the mature stage plays an important role in transporting
enormous amount of fine-grained sediments to the down streamside of China. Such transport of SS during this stage is
much higher than those during other stages and most of these sediments are supplied from the resuspended mudsediments
of the ECS, with origins mainly in Yangtze River. This study suggests that the resuspension and outflow of the
sediment plume is primarily caused by intensive mixing and existence of the coastal and offshore circulation features
during the mature stage of the SS evolution.
Geostationary Ocean Color Imager (GOCI) onboard its Communication Ocean and Meteorological Satellite (COMS) is scheduled for launch in 2008. GOCI includes the eight visible-to-near-infrared (NIR) bands, 0.5km pixel resolution, and a coverage region of 2500 x 2500km centered at 36N and 130E. GOCI has had the scope of its objectives broadened to understand the role of the oceans and ocean productivity in the climate system, biogeochemical variables, geological and biological response to physical dynamics and to detect and monitor toxic algal blooms of notable extension through observations of ocean color. To achieve these mission objectives, it is necessary to develop an atmospheric correction technique which is capable of delivering geophysical products, particularly for highly turbid coastal regions that are often dominated by strongly absorbing aerosols from the adjacent continental/desert areas. In this paper, we present a more realistic and cost-effective atmospheric correction method which takes into account the contribution of NIR radiances and include specialized models for strongly absorbing aerosols. This method was tested extensively on SeaWiFS ocean color imagery acquired over the Northwest Pacific waters. While the standard SeaWiFS atmospheric correction algorithm showed a pronounced overcorrection in the violet/blue or a complete failure in the presence of strongly absorbing aerosols (Asian dust or Yellow dust) over these regions, the new method was able to retrieve the water-leaving radiance and chlorophyll concentrations that were consistent with the in-situ observations. Such comparison demonstrated the efficiency of the new method in terms of removing the effects of highly absorbing aerosols and improving the accuracy of water-leaving radiance and chlorophyll retrievals with SeaWiFS imagery.
Accurate detection of highly toxic red tide algal blooms in coastal turbid waters has been challenging with currently existing spectral and bio-optical methods applied to satellite ocean color imagery, mainly because of the eventual interference of absorbing and scattering properties of dissolved organic and particulate inorganic matters with these methods. In the present study, we have presented a new red tide index (RI) technique to effectively identify the highly toxic dinoflagellate Cochlodinium polykrikoides (p) blooms in the Korean South Sea and neighboring waters. The effectiveness of this technique was evaluated using in-situ bio-optical observations and SeaWiFS ocean color imagery acquired during two bloom episodes on 19 September 2000 and 28 September 2003. The findings revealed that chlorophyll-a estimated through the application of OC-4 bio-optical algorithm to the SeaWiFS imagery falsely identified Cochlodinium.p blooms in areas abundance in colored dissolved organic and particulate inorganic matter constituents around coastal areas and river mouths. In contrast, red tide index was found to provide more accurate and comparable spatial Cochlodinium.p patterns consistent with in-situ observations, proving to be the best method for providing improved capability of detecting, predicting and monitoring of Cochlodinium.p bloom dynamics in clear oceanic waters and high scattering and absorbing waters off the Korean coast.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.