Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Morphological Detection and Classification of Microplastics and Nanoplastics Emerged from Consumer Products by Deep Learning

Hadi Rezvani, Navid Zarrabi, Ishaan Mehta, Christopher Kolios, Hussein Ali Jaafar, Cheng-Hao Kao, Sajad Saeedi, Nariman Yousefi

Abstract

Plastic pollution presents an escalating global issue, impacting health and environmental systems, with micro- and nanoplastics found across mediums from potable water to air. Traditional methods for studying these contaminants are labor-intensive and time-consuming, necessitating a shift towards more efficient technologies. In response, this paper introduces micro- and nanoplastics (MiNa), a novel and open-source dataset engineered for the automatic detection and classification of micro and nanoplastics using object detection algorithms. The dataset, comprising scanning electron microscopy images simulated under realistic aquatic conditions, categorizes plastics by polymer type across a broad size spectrum. We demonstrate the application of state-of-the-art detection algorithms on MiNa, assessing their effectiveness and identifying the unique challenges and potential of each method. The dataset not only fills a critical gap in available resources for microplastic research but also provides a robust foundation for future advancements in the field.

Morphological Detection and Classification of Microplastics and Nanoplastics Emerged from Consumer Products by Deep Learning

Abstract

Plastic pollution presents an escalating global issue, impacting health and environmental systems, with micro- and nanoplastics found across mediums from potable water to air. Traditional methods for studying these contaminants are labor-intensive and time-consuming, necessitating a shift towards more efficient technologies. In response, this paper introduces micro- and nanoplastics (MiNa), a novel and open-source dataset engineered for the automatic detection and classification of micro and nanoplastics using object detection algorithms. The dataset, comprising scanning electron microscopy images simulated under realistic aquatic conditions, categorizes plastics by polymer type across a broad size spectrum. We demonstrate the application of state-of-the-art detection algorithms on MiNa, assessing their effectiveness and identifying the unique challenges and potential of each method. The dataset not only fills a critical gap in available resources for microplastic research but also provides a robust foundation for future advancements in the field.
Paper Structure (21 sections, 4 equations, 13 figures, 6 tables)

This paper contains 21 sections, 4 equations, 13 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Sample Preparation Process
  4. Data Organization
  5. Data Annotation
  6. Data pre-processing
  7. SAM guided segmentation
  8. Manual review
  9. Automated review
  10. Data Statistics
  11. Data Analysis
  12. Evaluation protocol
  13. Configuration of the Dataset
  14. Metrics
  15. Networks
  16. ...and 6 more sections

Figures (13)

  • Figure 1: Polymer degradation and emergence experimental setup containing nine pieces of PS, PP, PET, and PE polymer in photoreactors
  • Figure 2: Microplastics images in the dataset
  • Figure 3: Size distribution of the emerged from PS, PP, PET, and PE in the course of 12 weeks
  • Figure 4: Results of detection using YOLOv10: (a) Network misses partial appearances of particles, (b) Overlapping particles are often overlooked, (c) Large particle size variance leads to missed instances, and (d) Small particles with low contrast are frequently missed.
  • Figure 5: Results of detection using Faster R-CNN: (a) Network detects partial appearances of particles in most cases. (b) Overlapping particles are handled better than YOLOv10. (c) A missed partial appearance of particles in a more complex image with different particle shapes and sizes. (d) Small particles are ignored less often compared to YOLOv10.
  • ...and 8 more figures