We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
YOLOv8-Based Microplastic Detection and Quantification in River Water Microscopic Images
AI summary Read the abstract
Researchers trained a YOLOv8 deep learning model on 300 microscopic images of river water from three Indonesian sites, achieving 0.786 precision and 0.731 mAP@0.5 for detecting microplastics, with higher image resolution improving accuracy and FTIR laboratory validation confirming compositional agreement with model detections.
Plastic particles with various size variations such as microplastics are environmental contaminants that are widely found in waters and have the potential to cause negative impacts. The process of identifying plastic particles using microscopic imagery manually takes a lot of time and considerable cost. In order to provide an alternative solution as part of early detection, microscopic image-based plastic particle detection was carried out with the YOLOv8 architecture, accompanied by an estimate of microplastic abundance in microplastic units per cubic meter. This study aims to develop and evaluate the detection of plastic particles in microscopic images of river water. This research dataset consists of 300 microscopic images taken from three river locations in Indonesia and annotated for model training and testing. The results of the evaluation showed that the proposed model had an aggregate performance value with a precision value of 0.786, recall of 0.66, and mAP@0.5 of 0.731. Additional test results show that with the addition of image resolution, the precision value can increase to 0.804 and the value mAP@0.5 increases to 0.762, even at the expense of computing time, which is also increasing. Extended scenario-based analysis showed that more than 87% of the detected objects fell into the category of small objects, affecting the localization sensitivity and variability of the estimated MPS value. This study also validated the results of object detection with FTIR-based laboratory tests using a full quantitative agreement between the model detection results and the identification of plastic particle materials at the sampling location level. The main contribution and findings of this study is an integrated evaluation framework for object detection, particle size characterization which is expected to be an alternative to the initial screening tool for plastic particle content.
More Papers Like This
Deep Learning Approaches for Detection and Classification of Microplastics in Water for Clean Water Management
AI summary Read the abstract
Researchers applied dual deep learning models (YOLOv8, YOLOv11, and several CNN architectures) to detect and classify microplastics in water, finding that these AI approaches could accurately identify plastic types across both aquatic and non-aquatic datasets.
Precise identification and validation of microplastics using deep learning models in the Yellow River
AI summary Read the abstract
Researchers tested computer vision models to detect microplastics in water samples from China's Yellow River, where muddy sediment makes identification difficult. One model combination worked best, accurately counting microplastics with results matching a standard lab identification method, offering a tool for monitoring microplastics in sediment-heavy rivers.
Automated Microplastic Detection in Environmental Samples Using YOLOv8 Nano: A Lightweight Approach for Edge Deploymen
AI summary Read the abstract
Researchers trained the lightweight YOLOv8 Nano model on over 3,200 environmental images to detect microplastics automatically, achieving 84% precision and exporting the model to TensorFlow Lite format for deployment on mobile and embedded devices in resource-constrained field settings.
Automated Microplastic Detection in Environmental Samples Using YOLOv8 Nano: A Lightweight Approach for Edge Deploymen
AI summary Read the abstract
Researchers trained the lightweight YOLOv8 Nano model on over 3,200 environmental images to detect microplastics automatically, achieving 84% precision and exporting the model to TensorFlow Lite format for deployment on mobile and embedded devices in resource-constrained field settings.
Efficient Microplastic Detection in Water Using ResNet50 and Fluorescence Imaging
AI summary Read the abstract
Researchers applied a ResNet50 deep learning model to fluorescence microscopy images of water samples, achieving high-accuracy classification of microplastics, demonstrating that deep learning can efficiently automate microplastic identification from microscopy data.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.