Use of unmanned aerial vehicles for efficient beach litter monitoring

Estimation of plastic marine debris volumes on beaches using unmanned aerial vehicles and image processing based on deep learning

Plastic marine debris (PMD) is of global concern. To help address this problem, a novel approach for estimating PMD volumes using a combination of unmanned aerial vehicle (UAV) surveys and image processing based on deep learning is proposed. A three-dimensional model and orthoscopic image of a beach, constructed via Structure from Motion software using UAV-derived data, enabled PMD volumes to be computed by edge detection through image processing. The accuracy of the method was verified by estimating the volumes of test debris placed on a beach in known sizes and shapes. The proposed approach shows potential for estimating PMD volumes with an error of <5%. Compared with subjective methods based on beach surveys, this approach can accurately, rapidly, and objectively calculate the PMD volume on a beach and can be used to improve the efficiency of beach surveys and identify beaches that need preferential cleaning.

Mapping marine litter with Unmanned Aerial Systems: A showcase comparison among manual image screening and machine learning techniques

Recent works have shown the feasibility of Unmanned Aerial Systems (UAS) for monitoring marine pollution. We provide a comparison among techniques to detect and map marine litter objects on an UAS-derived orthophoto of a sandy beach-dune system. Manual image screening technique allowed a detailed description of marine litter categories. Random forest classifier returned the best-automated detection rate (F-score 70%), while convolutional neural network performed slightly worse (F-score 60%) due to a higher number of false positive detections. We show that automatic methods allow faster and more frequent surveys, while still providing a reliable density map of the marine litter load. Image manual screening should be preferred when the characterization of marine litter type and material is required. Our analysis suggests that the use of UAS-derived orthophoto is appropriate to obtain a detailed geolocation of marine litter items, requires much less human effort and allows a wider area coverage.

A Deep Learning Model for Automatic Plastic Mapping Using Unmanned Aerial Vehicle (UAV) Data

Although plastic pollution is one of the most noteworthy environmental issues nowadays, there is still a knowledge gap in terms of monitoring the spatial distribution of plastics, which is needed to prevent its negative effects and to plan mitigation actions. Unmanned Aerial Vehicles (UAVs) can provide suitable data for mapping floating plastic, but most of the methods require visual interpretation and manual labeling. The main goals of this paper are to determine the suitability of deep learning algorithms for automatic floating plastic extraction from UAV orthophotos, testing the possibility of differentiating plastic types, and exploring the relationship between spatial resolution and detectable plastic size, in order to define a methodology for UAV surveys to map floating plastic. Two study areas and three datasets were used to train and validate the models. An end-to-end semantic segmentation algorithm based on U-Net architecture using the ResUNet50 provided the highest accuracy to map different plastic materials (F1-score: Oriented Polystyrene (OPS): 0.86; Nylon: 0.88; Polyethylene terephthalate (PET): 0.92; plastic (in general): 0.78), showing its ability to identify plastic types. The classification accuracy decreased with the decrease in spatial resolution, performing best on 4 mm resolution images for all kinds of plastic. The model provided reliable estimates of the area and volume of the plastics, which is crucial information for a cleaning campaign.

Unmanned Aerial Vehicles for Debris Survey in Coastal Areas: Long-Term Monitoring Programme to Study Spatial and Temporal Accumulation of the Dynamics of Beached Marine Litter

Unmanned aerial vehicles (UAVs) are becoming increasingly accessible tools with widespread use as environmental monitoring systems. They can be used for anthropogenic marine debris survey, a recently growing research field. In fact, while the increasing efforts for offshore investigations lead to a considerable collection of data on this type of pollution in the open sea, there is still little knowledge of the materials deposited along the coasts and the mechanism that leads to their accumulation pattern. UAVs can be effective in bridging this gap by increasing the amount of data acquired to study coastal deposits, while also limiting the anthropogenic impact in protected areas. In this study, UAVs have been used to acquire geo-referenced RGB images in a selected zone of a protected marine area (the Migliarino, Massacciuccoli, and San Rossore park near Pisa, Italy), during a long-term (ten months) monitoring programme. A post processing system based on visual interpretation of the images allows the localization and identification of the anthropogenic marine debris within the scanned area, and the estimation of their spatial and temporal distribution in different zones of the beach. These results provide an opportunity to investigate the dynamics of accumulation over time, suggesting that our approach might be appropriate for monitoring and collecting such data in isolated, and especially in protected, areas with significant benefits for different types of stakeholders.

Mapping marine litter using UAS on a beach-dune system: a multidisciplinary approach

The amount of marine litter, mainly composed by plastic materials, has become a global environmental issue in coastal environments. Traditional monitoring programs are based on in-situ visual census, which require human effort and are time-demanding. Therefore, it is crucial to implement innovative mapping strategies to improve the environmental monitoring of marine litter on the coast. This work presents a procedure for an automated Unmanned Aerial System (UAS)-based marine litter mapping on a beach-dune system. A multidisciplinary framework, which comprises photogrammetry, geomorphology, machine learning and hydrodynamic modelling, was developed to process a block of UAS images. The work shows how each of these scientific methodologies can be complementary to improve and making more efficient the mapping of marine litter items with UAS on coastal environment. The very high-resolution orthophoto produced from UAS images was automatically screened by random forest machine learning method, in order to characterize the marine litter load on beach and dune areas, distinctively. The marine litter objects were identified with a F-test score of 75% when compared to manual procedure. The location of major marine litter loads within the monitored area was found related to beach slope and water level dynamics on the beach profiles, suggesting that UAS flight deployment and post-processing for beach litter mapping can be optimized based on these environmental parameters. The described UAS-based marine litter detection framework is intended to support scientists, engineers and decision makers aiming at monitoring marine and coastal pollution, with the additional aim of optimizing and automating beach clean-up operations.

Field test of beach litter assessment by commercial aerial drone

The visual survey is the most common method to quantify and characterize beach litter. However, it is very labor intensive and difficult to carry out on beaches which are remote or difficult to access. We suggest an alternative approach for assessing beach litter using an unmanned aerial vehicle (UAV), or aerial drone, with automated image requisition and processing. Litter of different sizes, colours, and materials were placed randomly on two beaches. Images of beaches with different substrates were obtained by the drone at different operating heights and light conditions and litter on the beaches was identified from the photos by untrained personnel. The quantification of beach litter using the drone was three times faster than that by visual census. This study has demonstrated the potential of using the drone as a cost-effective and an efficient sampling method in routine beach litter monitoring programs.

Vessel-based photographic assessment of beach litter in remote coasts. A wide scale application in Saronikos Gulf, Greece

The abundance of marine debris was quantified for a total of sixty-two inaccessible beaches in the western Saronikos Gulf, Greece. High resolution images were obtained through vessel-based photography survey, merged into seamless photomosaics, and manually processed to quantify beach litter abundance. A sample of four selected beaches were subjected to detailed photography followed by beach macro-litter (≥ 2.5 cm) in-situ sampling surveys over a period of one year, to calibrate and validate the proposed method. Regression analysis between photographic and in-situ data showed a significant correlation, hence providing a highly accurate regression model to assess the real number of beach litter stranded on the rest of the investigated beaches, exhibiting clear correlations to the hydrodynamic status of the area and, provide an indication of the main litter sources. The proposed method is an easily applicable and useful tool for fast and low-cost macro-litter monitoring in extended, remote coastlines, when only photographic data are available.

Anthropogenic Marine Debris assessment with Unmanned Aerial Vehicle imagery and deep learning: A case study along the beaches of the Republic of Maldives

Anthropogenic Marine Debris (AMD) is one of the major environmental issues of our planet to date, and plastic accounts for 80% of total AMD. Beaches represent one of the main marine compartment where AMD accumulates, but few and scattered regional assessments are available from literature reporting quantitative estimation of AMD distributed on the shorelines. However, accessing information on the AMD accumulation rate on beaches, and the associated spatiotemporal oscillations, would be crucial to refining global estimation on the dispersal mechanisms. In our work, we address this issue by proposing an ad-hoc methodology for monitoring and automatically quantifying AMD, based on the combined use of a commercial Unmanned Aerial Vehicle (UAV) (equipped with an RGB high-resolution camera) and a deep-learning based software (i.e.: PlasticFinder). Remote areas were monitored by UAV and were inspected by operators on the ground to check and to categorise all AMD dispersed on the beach. The high-resolution images obtained from UAV allowed to visually detect a percentage of the objects on the shores higher than 87.8%, thus providing suitable images to populate training and testing datasets, as well as gold standards to evaluate the software performance. PlasticFinder reached a Sensitivity of 67%, with a Positive Predictive Value of 94%, in the automatic detection of AMD, but a limitation was found, due to reduced sunlight conditions, thus restricting to the use of the software in its present version. We, therefore, confirmed the efficiency of commercial UAVs as tools for AMD monitoring and demonstrated - for the first time - the potential of deep learning for the automatic detection and quantification of AMD.

Seasonality of marine plastic abundance in central Red Sea pelagic waters

The Red Sea holds one of the lowest concentrations of floating plastic worldwide and no evident congregation zones were identified so far, despite peculiar oceanographic features that candidate the basin as an accumulation area for floating debris. However, the Red Sea exhibits a complex pattern of surface currents, which changes according to the monsoon season, possibly affecting the abundance of plastic throughout the year. To explore the effect of seasonality on plastic concentration in surface waters, we conducted a fortnightly time series sampling, using a neuston net, for 21 months at a pelagic station in the central Red Sea, where the major seasonal overturn of the Red Sea surface circulation occurs. The estimated average abundance (±SE) was 58,563 ± 19,272 items Km^-2 (73.5 ± 40.75 g Km^-2), highly variable according to season, lower during the summer monsoon. Indeed, the winter monsoon pushes oceanic surface waters inside the Red Sea, transporting alongside floating plastic items, whilst surface currents exit the basin during the summer monsoon, depleting central Red Sea waters from floating plastic. Composition of plastic items also changes through time. Particularly, the higher proportion of films and foams during summer months suggests that the main source of plastic at the sampling station from June to September is a short-range transport, while during winter months, the higher contribution of small fragments indicates that, from October to May, plastic is also transported to the central Red Sea through surface currents for long distances, possibly coming all the way from the Indian Ocean.

Seasonality of riverine macroplastic transport

Marine plastic pollution is an increasing environmental threat. Although it is assumed that most marine plastics are transported from land to the ocean through rivers, only limited data on riverine plastic transport exists. Recently, new methods have been introduced to characterize riverine plastics consistently through time and space. For example, combining visual counting observations and plastic debris sampling can provide order of magnitude estimations of plastic transport through a river. In this paper, we present findings from multi-season measurement campaign in the Saigon River, Vietnam. For the first time, we demonstrate that macroplastic transport exhibits strong temporal variation. The monthly averaged plastic transport changes up to a factor five within the measurement period. As it is unclear what drives the variation in plastic transport, relations between rainfall, river discharge, presence of organic material and plastic transport have been explored. Furthermore, we present new findings on the cross-sectional and vertical distribution of riverine plastic transport. With this paper we present new insights in the origin and fate of riverine plastic transport, emphasizing the severity of the emerging thread of plastic pollution on riverine ecosystems.

Microplastic removal by Red Sea giant clam (Tridacna maxima)

This study assesses for the first time the ingestion of microplastics by giant clams and evaluates their importance as a sink for this pollutant. A total of 24 individuals of two size classes were collected from the Red Sea and then exposed for 12 days to 4 concentrations of polyethylene microbeads ranging from 53 to 500 μm. Experiments revealed that clams actively take up microplastic from the water column and the average of beads retained inside the animal was ∼7.55 ± 1.89 beads individual ^-1 day ^-1 (5.76 ± 1.16 MPs/g dw). However, the digestive tract itself cannot be considered the only sink of microbeads in Tridacnids. Indeed, shells play a key role as well. The abundance of microplastic adhering to the shells, which was estimated directly, was positively correlated to the concentration of beads found in the surrounding seawater. Therefore, clams' shells contribute to the removal of 66.03 ± 2.50% of the microplastic present in the water column. Furthermore, stress responses to the exposure to polyethylene were investigated. Gross Primary Production:Respiration (GPP:R) ratio decreased throughout of the experiment, but no significant difference was found between treatments and controls.

Riverine Plastic Litter Monitoring Using Unmanned Aerial Vehicles (UAVs)

Plastic debris has become an abundant pollutant in marine, coastal and riverine environments, posing a large threat to aquatic life. Effective measures to mitigate and prevent marine plastic pollution require a thorough understanding of its origin and eventual fate. Several models have estimated that land-based sources are the main source of marine plastic pollution, although field data to substantiate these estimates remain limited. Current methodologies to measure riverine plastic transport require the availability of infrastructure and accessible riverbanks, but, to obtain measurements on a higher spatial and temporal scale, new monitoring methods are required. This paper presents a new methodology for quantifying riverine plastic debris using Unmanned Aerial Vehicles (UAVs), including a first application on Klang River, Malaysia. Additional plastic measurements were done in parallel with the UAV-based approach to make comparisons between the two methods. The spatiotemporal distribution of the plastics obtained with both methods show similar patterns and variations. With this, we show that UAV-based monitoring methods are a promising alternative for currently available approaches for monitoring riverine plastic transport, especially in remote and inaccessible areas.

Floating Xylene Spill Segmentation from Ultraviolet Images via Target Enhancement

Automatic colorless floating hazardous and noxious substances (HNS) spill segmentation is an emerging research topic. Xylene is one of the priority HNSs since it poses a high risk of being involved in an HNS incident. This paper presents a novel algorithm for the target enhancement of xylene spills and their segmentation in ultraviolet (UV) images. To improve the contrast between targets and backgrounds (waves, sun reflections, and shadows), we developed a global background suppression (GBS) method to remove the irrelevant objects from the background, which is followed by an adaptive target enhancement (ATE) method to enhance the target. Based on the histogram information of the processed image, we designed an automatic algorithm to calculate the optimal number of clusters, which is usually manually determined in traditional cluster segmentation methods. In addition, necessary pre-segmentation processing and post-segmentation processing were adopted in order to improve the performance. Experimental results on our UV image datasets demonstrated that the proposed method can achieve good segmentation results for chemical spills from different backgrounds, especially for images with strong waves, uneven intensities, and low contrast.

Mangrove forests as traps for marine litter

To verify weather mangroves act as sinks for marine litter, we surveyed through visual census 20 forests along the Red Sea and the Arabian Gulf, both in inhabited and remote locations. Anthropogenic debris items were counted and classified along transects, and the influence of main drivers of distribution were considered (i.e. land-based and ocean-based sources, density of the forest and properties of the object). We confirmed that distance to major maritime traffic routes significantly affects the density of anthropogenic debris in Red Sea mangrove forests, while this was independent of land-based activities. This suggests ocean-based activities combined with surface currents as major drivers of litter in this basin. Additionally, litter was more abundant where the mangrove density was higher, and object distribution through the mangrove stand often depended on their shape and dimension. We particularly show that pneumatophores act as a sieve retaining large plastic objects, leading to higher plastic mass estimates in mangroves compared to those of beaches previously surveyed in the Red Sea.

A novel GIS-based tool for predicting coastal litter accumulation and optimising coastal cleanup actions

Effective site selection is a key component of maximising debris removal during coastal cleanup actions. We tested a GIS-based predictive model to identify marine litter hotspots in Lofoten, Norway based on shoreline gradient and shape. Litter density was recorded at 27 randomly selected locations with 5 transects sampled in each. Shoreline gradient was a limiting factor to litter accumulation when >35%. The curvature of the coastline correlated differently with litter density at different spatial scales. The greatest litter concentrations were in small coves located on larger headlands. A parsimonious model scoring sites on a scale of 1-5 based on shoreline slope and shape had the highest validation success. Sites unlikely to have high litter concentrations were successfully identified and could be avoided. The accuracy of hotspot identifications was more variable, and presumably more parameters influencing litter deposition, such as shoreline aspect relative to prevailing winds, should be incorporated.

Investigating the Utility Potential of Low-Cost Unmanned Aerial Vehicles in the Temporal Monitoring of a Landfill

The collection of solid waste is a challenging issue, especially in highly urbanized areas. In developing countries, landfilling is currently the preferred method for disposing of solid waste, but each landfill has a limited lifecycle. Therefore, changes in the amount of stored waste should be monitored for the sustainable management of such areas. In this study, volumetric changes in a landfill were examined using a low-cost unmanned aerial vehicle (UAV). Aerial photographs obtained from five different flights, covering approximately two years, were used in the volume calculations. Values representing the amount of remaining space between the solid waste and a reference plane were determined using digital elevation models, which were produced based on the structure from motion (SfM) approach. The obtained results and potential of UAVs in the photogrammetric survey of a landfill were further evaluated and interpreted by considering other possible techniques, ongoing progress, and the information existing in an environmental impact assessment report. As a result of the study, it was proved that SfM carried out using a low-cost UAV has a high potential for use in the reconstruction of a landfill. Outcomes were obtained over a short period, without the need for direct contact with the solid waste, making the UAV preferable for use in planning and decision-making studies.

Drones for Conservation in Protected Areas: Present and Future

Park managers call for cost-effective and innovative solutions to handle a wide variety of environmental problems that threaten biodiversity in protected areas. Recently, drones have been called upon to revolutionize conservation and hold great potential to evolve and raise better-informed decisions to assist management. Despite great expectations, the benefits that drones could bring to foster effectiveness remain fundamentally unexplored. To address this gap, we performed a literature review about the use of drones in conservation. We selected a total of 256 studies, of which 99 were carried out in protected areas. We classified the studies in five distinct areas of applications: “wildlife monitoring and management”; “ecosystem monitoring”; “law enforcement”; “ecotourism”; and “environmental management and disaster response”. We also identified specific gaps and challenges that would allow for the expansion of critical research or monitoring. Our results support the evidence that drones hold merits to serve conservation actions and reinforce effective management, but multidisciplinary research must resolve the operational and analytical shortcomings that undermine the prospects for drones integration in protected areas.

Proof of concept for a model of light reflectance of plastics floating on natural waters

Remote sensing of plastic littering natural waters is an emerging field of science with the potential to provide observations on local to global scales. We present the verification of a theoretical reflectance model of sunlight interacting with a water surface littered with buoyant plastic objects. We measured a few common litter items of different polymers as well as shapes, transparencies, and surface roughnesses. Spectral reflectance measurements in the field were backed up with measurements in the laboratory of coefficients of total and diffuse reflectance, transmittance and absorption. We evaluated a single-band algorithm for 850 nm wavelength and a dual-band algorithm using a second wavelength at a polymer absorption band between 1660 and 1730 nm. Both algorithms were plastic litter type specific. Our findings show that for interpreting spectral remote sensing of floating plastic, physical properties that control geometrical optics should complement information about the absorption spectra of the polymer.