Abstract
Pollution in the form of litter in the natural environment is one of the great challenges of our times. Automated litter detection can help assess waste occurrences in the environment. Different machine learning solutions have been explored to develop litter detection tools, thereby supporting research, citizen science, and volunteer clean-up initiatives. However, to the best of our knowledge, no work has investigated the performance of state-of-the-art deep learning object detection approaches in the context of litter detection. In particular, no studies have focused on the assessment of those methods aiming their use in devices with low processing capabilities, e.g., mobile phones, typically employed in citizen science activities. In this paper, we fill this literature gap. We performed a comparative study involving state-of-the-art CNN architectures (e.g., Faster RCNN, Mask-RCNN, EfficientDet, RetinaNet and YOLO-v5), two litter image datasets and a smartphone. We also introduce a new dataset for litter detection, named PlastOPol, composed of 2418 images and 5300 annotations. The experimental results demonstrate that object detectors based on the YOLO family are promising for the construction of litter detection solutions, with superior performance in terms of detection accuracy, processing time, and memory footprint.
Highlights
Nowadays, litter is one of the greatest challenges [1], due to its ubiquity, nature and scale, with dire consequences for freshwater and marine ecosystems [2,3,4,5], and, arguably, for urban environments and human health
We present a new dataset, named “PlastOPol,” which is based on images taken through the Marine Debris Tracker with the goal of giving, to the computer science and environmental communities, a new set of 2418 images with the presence of litter in a realistic context covering several types of environments, i.e., urban, beaches, forests and flint fields, and including different types of litter, including plastic, glass, metal, paper, cloth and rubber, among others
We discuss the effectiveness and efficiency of two compact neural networks (YOLO-v5s [32] and EfficientDet-d0 [30]) on a mobile set-up (Section 5.4). We considered these tiny deep neural networks with the goal of evaluating the trade-off between effectiveness and efficiency when running these approaches on devices with computational constraints
Summary
Litter is one of the greatest challenges [1], due to its ubiquity, nature and scale, with dire consequences for freshwater and marine ecosystems [2,3,4,5], and, arguably, for urban environments and human health. Plastic items can cause lethal damage to aquatic and terrestrial species through entanglement, gut perforation, and starvation [6]. Plastic decomposes, creates microplastics and, nanoplastics which enter the food chain [7]. To investigate and tackle this challenge, decision-makers require reliable data— on the sources of waste, and on its composition, distribution and magnitude over large geographic areas. One way of providing technological support is to automate the process of litter logging and litter detection, especially in areas of difficult access, such as forest and mountains, so as to make the process more effective for volunteers and researchers and enable the processing of large datasets, while minimizing resources, fatigue, and even risks
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
