High-Efficiency Method for Estimating Waste Coverage at River Mouths

Marine debris has become a critical global environmental issue, and river mouths are one of the primary pathways through which land-based waste enters the oceans. Therefore, accurate quantification of marine debris at river mouths is essential for coastal management and marine-conservation-related policymaking. However, existing survey-based approaches largely rely on manual visual inspection, which is not only time-consuming but also subject to human bias; thus, these approaches cannot be used to effectively assess marine debris coverage across large areas. Some studies have applied artificial-intelligence-based image recognition techniques to assess marine debris coverage, but most of these studies have focused on debris classification rather than quantitative estimation of debris coverage. To address these limitations, the present study developed an efficient and accurate method for estimating the debris coverage at river mouths, thereby improving upon the subjectivity and inefficiency inherent in traditional visual survey methods. Image data from two sources were used in this study: (1) images captured under various simulated scenarios by high-resolution cameras and (2) high-resolution aerial images acquired by unmanned aerial vehicles (UAVs) at river mouths. Following image acquisition, an artificial-intelligence-based image analysis system was employed to perform preprocessing procedures on the images—including noise reduction, grayscale conversion, binarization, and edge detection—to quantify the proportion of the debris-covered area within each image. To validate the reliability of the image-based estimates, this study adopted an aerial grid method as a reference benchmark. Aerial grids were overlaid on images of the examined areas, and the proportion of debris within each grid cell was manually calculated to determine the actual debris coverage. The accuracy of the proposed methodology was evaluated by comparing its results with those derived using the aerial grid method, and potential sources of error were examined. The results indicated that integration of UAV aerial imaging, image processing techniques, and coverage quantification methods enables the feasible and accurate estimation of the debris coverage at river mouths. The proposed approach can assist governmental agencies and nongovernmental organizations in the evidence-based planning of marine debris monitoring programs, marine debris cleanup efforts, and relevant conservation policies.