We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
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.
Investigating the spatial distribution and seasonal variations of riverine microplastics (MPs) is vital to aquatic ecological research. Mass suspended sediment within the Yellow River introduces severe background noise, which impedes conventional deep learning models from accurately extracting fine edge features of microplastics. Based on water samples collected from the Zhengzhou reach of the Yellow River, this study systematically evaluates a matrix of 24 hybrid encoder-decoder segmentation architectures covering six backbones (ResNet and ResNeXt series) and four decoders (FPN, DeepLabV3+, UNet, UNet++). The objective is to identify the optimal single model configuration (including network depth, topological layout, and feature fusion strategy) capable of accurately extracting fine edge features of microplastics under severe sediment interference.. Beyond standard segmentation metrics, three dedicated indices including Counting Error (CE), Background Complexity (BC) and 95% Hausdorff Distance (HD95) are proposed to quantify the anti-noise robustness of each model. Grad-CAM visualization is adopted to interpret model spatial feature activation and target focus patterns, and micro-Fourier transform infrared (μ-FTIR) spectroscopy is utilized to cross-validate segmentation outputs. The ResNeXt-101 + UNet++ hybrid network delivers optimal performance. Equipped with grouped convolution and nested dense skip connections, this architecture bridges the information gap between shallow spatial details and deep semantic features, achieving a precision of 0.9534, a recall of 0.9991 and a mean absolute counting error (MAE) of 0.242. Microplastic abundances predicted by the model show excellent consistency with μ-FTIR measurements, addressing the common disconnection between image segmentation and spectroscopic identification in existing studies. This work develops an accurate intelligent pipeline for microplastic detection and quantification, facilitating routine surveillance and ecological risk assessment within the Yellow River Basin, while providing a methodological framework that warrants further validation in other sediment-laden aquatic environments.
More Papers Like This
Efficient and accurate microplastics identification and segmentation in urban waters using convolutional neural networks
AI summary Read the abstract
Researchers developed convolutional neural network models for efficiently identifying and segmenting microplastics in urban water samples from southern China. The study found that deep learning approaches can significantly reduce the time and labor required for microplastic identification compared to manual methods, offering a scalable tool for monitoring microplastic pollution in urban waterways.
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.
Automatic Detection of Microplastics in the Aqueous Environment
AI summary Read the abstract
Researchers developed a deep-learning system for real-time detection and counting of microplastics in freshwater, achieving high accuracy for particles 1 mm and larger.
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.
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.