Enhancing Georeferencing and Mosaicking Techniques over Water Surfaces with High-Resolution Unmanned Aerial Vehicle (UAV) Imagery

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(2), P. 290 - 290

Published: Jan. 11, 2024

Aquatic ecosystems are crucial in preserving biodiversity, regulating biogeochemical cycles, and sustaining human life; however, their resilience against climate change anthropogenic stressors remains poorly understood. Recently, unmanned aerial vehicles (UAVs) have become a vital monitoring tool, bridging the gap between satellite imagery ground-based observations coastal marine environments with high spatial resolution. The dynamic nature of water surfaces poses challenge for photogrammetric techniques due to absence fixed reference points. Addressing these issues, this study introduces an innovative, efficient, accurate workflow georeferencing mosaicking that overcomes previous limitations. Using open-source Python libraries, employs direct produce georeferenced orthomosaic integrates multiple UAV captures, has been tested locations worldwide optical RGB, thermal, multispectral imagery. best case achieved Root Mean Square Error 4.52 m standard deviation 2.51 accuracy, thus UAV’s centimeter-scale This represents significant advancement processes, resolving major limitation facing technology remote observation local-scale phenomena over surfaces.

Language: Английский

---

Monitoring the marine invasive alien species Rugulopteryx okamurae using unmanned aerial vehicles and satellites

Frontiers in Marine Science, Journal Year: 2022, Volume and Issue: 9

Published: Oct. 13, 2022

Rugulopteryx okamurae is a species of brown macroalgae belonging to the Dictyotaceae family and native north-western Pacific. As an Invasive Alien Species (IAS), it was first detected in Strait Gibraltar 2015. Since then, R. has been spreading rapidly through submerged euphotic zone, colonizing from 0 50 m depth generating substantial economic environmental impacts on Andalusian coasts (southern Spain). More than 40% marine IAS European Union (EU) are macroalgae, representing one main threats biodiversity ecosystem functioning coastal habitats. This study presents monitoring pilot beached fresh down 5 Tarifa (Cadiz, Spain), combining multispectral remote sensing data collected by sensors on-board Unmanned Aerial Vehicles (UAVs) satellites, how this information can be used support decision-making policy. We UAV flight carried out at Bolonia beach (Tarifa, Spain) 1 st July 2021 Sentinel-2 (S2) Landsat-8 (L8) image acquisitions close drone date. In situ were also measured same date flight, they train supervised classification Super Vector Machine (SVM) method based spectral obtained for each substrate cover. The results show images allow detection , accuracy water, land vegetation, sand depending resolution (8.3 cm/pixel 10 m/pixel S2 30 L8). While imagery precisely delimited area occupied satellite capable detecting its presence, able generate early warnings. demonstrates usefulness techniques incorporated continuous programmes areas. key supporting regional, national policies order adapt strategic management invasive macrophytes.

Language: Английский

---

Fusion of Drone-Based RGB and Multi-Spectral Imagery for Shallow Water Bathymetry Inversion

Remote Sensing, Journal Year: 2022, Volume and Issue: 14(5), P. 1127 - 1127

Published: Feb. 24, 2022

Shallow bathymetry inversion algorithms have long been applied in various types of remote sensing imagery with relative success. However, this approach requires that increased radiometric resolution the visible spectrum be available. The recent developments drones and camera sensors allow for testing current techniques on new datasets centimeter resolution. This study explores bathymetric mapping capabilities fused RGB multispectral as an alternative to costly hyperspectral drones. Combining drone-based into a single cube dataset provides necessary detail shallow applications. technique is based commercial open-source software does not require input reference depth measurements contrast other approaches. robustness method was tested three different coastal sites contrasting seafloor maximum six meters. use suitable end-member spectra, which are representative area, important parameters model tuning. results promising, showing good correlation (R2 > 0.75 Lin’s coefficient 0.80) less than half meter average error when they compared sonar measurements. Consequently, integration from (visible range) assists producing detailed maps small-scale areas optical modelling.

Language: Английский

---

Drone-Based Characterization of Seagrass Habitats in the Tropical Waters of Zanzibar

Remote Sensing, Journal Year: 2022, Volume and Issue: 14(3), P. 680 - 680

Published: Jan. 31, 2022

Unmanned automatic systems (UAS) are increasingly being applied as an alternative to more costly time-consuming traditional methods for mapping and monitoring marine shallow-water ecosystems. Here, we demonstrate the utility of combining aerial drones with in situ imagery characterize habitat conditions nine seagrass-dominated areas on Unguja Island, Zanzibar. We object-based image analysis a maximum likelihood algorithm drone images derive cover maps important seagrass parameters: composition; species; horizontal- depth-percent covers, seascape fragmentation. mapped sites covering 724 ha, categorized into seagrasses (55%), bare sediment (31%), corals (9%), macroalgae (5%). An average six species were found, 20% “dense cover” (40–70%). achieved high map accuracy types (87%), (80%), (76%). In all sites, observed clear decreases covers depths ranging from 30% at 1–2 m, 1.6% 4–5 m depth. The depth dependency varied significantly among species. Areas associated low also had fragmented distribution pattern, scattered populations. was correlated negatively (r2 = 0.9, p < 0.01) sea urchins. A multivariate similarity (ANOSIM) biotic features, derived data, suggested that could be organized three different coastal types. This study demonstrates robustness characterizing complex tropical waters. recommend adopting drones, combined photos, establishing suite data relevant ecosystem Western Indian Ocean (WIO).

Language: Английский

---

Precision Aquaculture Drone Mapping of the Spatial Distribution of Kappaphycus alvarezii Biomass and Carrageenan

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(14), P. 3674 - 3674

Published: July 23, 2023

The aquaculture of Kappaphycus alvarezii (Kappaphycus hereafter) seaweed has rapidly expanded among coastal communities in Indonesia due to its relatively simple farming process, low capital costs and short production cycles. This species is mainly cultivated for carrageenan content used as a gelling agent the food industry. To further assist producers improving cultivation management providing quantitative information about yield, novel approach involving remote sensing techniques was tested. In this study, multispectral images obtained from drone (Unoccupied Aerial Vehicle, UAV) were processed estimate fresh weights at site South Sulawesi. UAV imagery geometrically radiometrically corrected, resulting orthomosaics detecting classifying using random forest algorithm. classification results combined with situ measurements weight empirical relations between area seaweed/carrageenan. allowed quantifying biometry biochemistry single lines plot scales. Fresh estimated different dates within three distinct cycles, daily growth rate each cycle derived. Data upscaled small family-scale farm large-scale leader compared previous estimations. our knowledge, study provides, first time, an estimation yield scale by exploiting very high spatial resolution data. Overall, use proved be promising monitoring, opening way precision Kappaphycus.

Language: Английский

---

Species level mapping of a seagrass bed using an unmanned aerial vehicle and deep learning technique

PeerJ, Journal Year: 2022, Volume and Issue: 10, P. e14017 - e14017

Published: Oct. 17, 2022

Background Seagrass beds are essential habitats in coastal ecosystems, providing valuable ecosystem services, but threatened by various climate change and human activities. monitoring remote sensing have been conducted over past decades using satellite aerial images, which low resolution to analyze changes the composition of different seagrass species meadows. Recently, unmanned vehicles (UAVs) allowed us obtain much higher is promising observing fine-scale composition. Furthermore, image processing techniques based on deep learning can be applied discrimination that were difficult only color variation. In this study, we mapping a multispecific bed Saroma-ko Lagoon, Hokkaido, Japan, compared accuracy three methods areas composition, i.e ., pixel-based classification, object-based application neural network. Methods We set five benthic classes, two ( Zostera marina Z. japonica ), brown green macroalgae, no vegetation for creating cover map. High-resolution images UAV photography enabled produce map at fine scales (<1 cm resolution). Results The network successfully classified species. classification was highest (82%) when applied. Conclusion Our results highlighted combination could help monitor spatial extent classify their very scales.

Language: Английский

---

Low-cost UAV monitoring: insights into seasonal volumetric changes of an oyster reef in the German Wadden Sea

Frontiers in Marine Science, Journal Year: 2023, Volume and Issue: 10

Published: Oct. 9, 2023

This study aims to quantify the dimensions of an oyster reef over two years via low-cost unoccupied aerial vehicle (UAV) monitoring and examine seasonal volumetric changes. No current investigated UAV changes reef-building Pacific ( Magallana gigas ) in German Wadden Sea, considering uncertainty measurements processing. Previous studies have concentrated on classifying mapping smaller reefs using terrestrial laser scanning (TLS) or hyperspectral remote sensing data recorded by UAVs satellites. employed a consumer-grade with low spectral resolution semi-annually record for generating digital elevation models (DEM) orthomosaics structure from motion (SfM), enabling identifying oysters. The machine learning algorithm Random Forest (RF) proved be accurate classifier identify oysters low-spectral data. Based classified data, was spatially analysed, difference (DoDs) were used estimate introduction propagation errors supported determining vertical confidence level 68% 95%, highlighting significant change detection. results indicate volume increase 22 m³ loss 2 period, 95%. In particular, lost area between September 2020 March 2021, when exposed air more than ten hours. top increased -15.5 ± 3.6 cm NHN -14.8 3.9 2022, but could not determine consistent annual growth rate. As long as environmental hydrodynamic conditions are given, is expected continue growing higher elevations tidal flats, only limited exposure. rates suggest further expansion, resulting roughness surface that contributes flow damping altering sedimentation processes. Further proposed investigate limiting stressors, providing robust evidence regarding influence exposure loss.

Language: Английский

---

Subtidal seagrass and blue carbon mapping at the regional scale: a cloud-native multi-temporal Earth Observation approach

GIScience & Remote Sensing, Journal Year: 2024, Volume and Issue: 62(1)

Published: Dec. 16, 2024

The seagrass ecosystems are among the most important organic carbon sinks on Earth, having a key role as climate change buffers. Among all seagrasses, Posidonia oceanica, an endemic species in Mediterranean Sea, has been observed to feature highest stock and sequestration rate seagrasses. We developed satellite-based workflow complement situ monitoring efforts Balearic Islands (Western Mediterranean), reducing field expenses while covering regional spatial scales. Our synoptic tool uses Sentinel-2 A/B satellite imagery at 10 m resolution generate multi-temporal composite (2016–2022) of Islands' coastal waters within Google Earth Engine cloud computing platform, optimizing image processing highlighting importance high-resolution bathymetric dataset increase mapping accuracies. Machine learning algorithms have applied perform detection, obtaining cartography up 30 depth, estimating 505.6 km2 habitat extent. Using existing soil (Cstock) data, we estimated mean Cstock value 12.27 ± 2.1 million megagram (Mg) Corg, total annual C fixation (Cfix) (Cseq) rates P. oceanica 1,116.3 Mg Corg 227 according depth. methodology highlights using large archive optical optimized bathymetry better map account blue across showing integrate this Observation approach ensure ecosystem This information aims support development strategies with time- cost-efficient Sea.

Language: Английский

---

Mapping intertidal oyster farms using unmanned aerial vehicles (UAV) high-resolution multispectral data

Estuarine Coastal and Shelf Science, Journal Year: 2023, Volume and Issue: 291, P. 108432 - 108432

Published: July 11, 2023

In France, oyster aquaculture has been historically developed in intertidal zones, with shellfish farming areas covering much of the Atlantic coast. Monitoring these off-bottom cultures where oysters are grown plastic mesh-bags set on trestle tables is mandatory for maritime administration to check compliance a Structural Plan Document (SPD), while also being important stock assessment relation carrying capacity issues. However, traditional monitoring methods time-consuming, labor-intensive, and inefficient large areas. this study, we used new GIS-based analytical method assess potential high-resolution Unmanned Aerial Vehicle (UAV) multispectral data retrieve spatial information oyster-farming structures using Bourgneuf Bay (France) as case-study. A non-parametric machine learning algorithm was applied four UAV flight orthomosaics collected at different altitudes (12, 30, 50, 120 m) identify mesh-bags. These supervised classifications achieved overall accuracies above 95% all tested altitudes. addition, an accurate distinction oyster-bag mesh sizes (4, 9 14 mm) obtained 12–50 m flights, but there lower accuracy m. Across 4 mm size least well detected (72.14% Producer Accuracy). This can be bags specific mesh-sizes spat or adult grow out. Finally, accurately measured table heights Digital Surface Model (DSM) derived from Structure Motion (SfM) photogrammetry. The 50 suggested best compromise obtain precise measurements larger than altitude flights. demonstrates that technology provide variables relevant farmers coastal managers efficient, rapid, non-destructive way monitor extent characteristics regularly.

Language: Английский

---

High-spatial resolution UAV multispectral data complementing satellite imagery to characterize a chinstrap penguin colony ecosystem on deception island (Antarctica)

GIScience & Remote Sensing, Journal Year: 2022, Volume and Issue: 59(1), P. 1159 - 1176

Published: Aug. 5, 2022

Remote sensing has evolved as an alternative to traditional techniques in the spatio-temporal monitoring of Antarctic ecosystem, especially with rapid expansion use Unmanned Aerial Vehicles (UAVs), providing a centimeter-scale spatial resolution. In this study, potential high-spatial resolution multispectral sensor embedded UAV is compared medium satellite remote (Sentinel-2 and Landsat 8) monitor characteristics Vapor Col Chinstrap penguin (Pygoscelis antarcticus) colony ecosystem (Deception Island, South Shetlands Islands, Antarctica). Our main objective generate precise thematic maps typical colonies derived from supervised analysis spectral information obtained these sensors. For this, two parametric classification algorithms (Maximum Likelihood, MLC, Spectral Angle, SAC) non-parametric machine learning classifiers (Support Vector Machine, SVM, Random Forest, RFC) are tested imagery, obtaining best results SVM classifier (93.19% OA). study shows that outperforms imagery (87.26% OA Sentinel-2 Level 2 (S2L2) 70.77% 8 (L8L2) classification) characterization substrate due higher resolution, although differences between S2L2 minimal. Thus, both sensors used tandem could provide broader more view how area covered by different elements ecosystems can change over time global climate scenario. addition, represents takes place colony, estimating total coverage approximately 20,000 m2 guano areas period.

Language: Английский

---