We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Toward reliable rapid morphological screening: A systematic benchmark of deep learning models for microplastic detection in mangrove filter images
AI summary Read the abstract
Scientists trained an AI model to spot microplastics in water and soil samples in seconds instead of hours, matching results from slower lab tests with 92% accuracy. This could speed up microplastic monitoring worldwide, helping researchers track this pollution faster so we can better understand its risks to ecosystems and human health.
Microplastics are a global concern, but reliable high-throughput morphological candidate screening remains constrained by labor-intensive visual inspection and the limited availability of chemically supported image benchmarks. Here, we constructed an FTIR-supported image dataset from mangrove water and sediment samples and benchmarked 11 deep-learning detector configurations. The dataset contained 2865 unique FTIR-confirmed particles assigned to fragment, fiber, and film categories, together with background-only image tiles representing heterogeneous membrane residues and non-plastic interference. Among the evaluated configurations, YOLOv5m achieved the highest performance (F1 = 0.906; mAP@0.5 = 0.910). Model-derived particle counts showed a strong linear association with FTIR-confirmed counts (R = 0.92). Across 45 membranes, automated complete-image screening and counting with YOLOv5m required an average of 6.57 s per membrane. Performance was higher for in-distribution mangrove samples than for out-of-distribution samples, indicating that environmental background and particle heterogeneity remain important determinants of model transferability. Class-specific analysis further showed that fiber-related errors were dominated by missed detections rather than confusion with other morphological categories. The framework is intended as a morphological screening and triage tool before FTIR or Raman confirmation, providing a quick solution for microplastic monitoring in complex field samples.
More Papers Like This
A new approach to classifying polymer type of microplastics based on Faster-RCNN-FPN and spectroscopic imagery under ultraviolet light
AI summary Read the abstract
Scientists developed an AI-based method using UV light photography to automatically identify and classify different types of microplastics, achieving 86-88% accuracy. This approach is faster and cheaper than traditional lab analysis methods that require expensive equipment. Better detection tools like this are essential for understanding how widespread microplastic contamination really is in coastal environments where people live and eat seafood.
Computer vision-based microplastic identification in water and analysis of environmental microbial mechanisms
AI summary Read the abstract
Scientists built an AI tool that can spot microplastics in water samples with over 97% accuracy, making it much faster than current lab methods for tracking this pollution. The tool also revealed that different types of microplastics attract different communities of bacteria, which matters because these plastic particles can act as rafts for microbes (including potentially harmful ones) as they move through waterways we depend on for drinking water and food. Faster, cheaper monitoring like this could help researchers and regulators keep closer tabs on microplastic contamination before it reaches our tap water or seafood.
Microplastic Contamination Detection in Water using Deep Learning Models
AI summary Read the abstract
Scientists trained a computer program (using AI image recognition) to automatically spot tiny plastic particles in water samples, and it correctly identified microplastics 99.5% of the time—faster and easier than the manual microscope checks experts currently rely on. This matters because microplastics in our water supply are linked to potential health risks, so having a quick, accurate way to detect them could help water treatment facilities catch contamination sooner and keep drinking water safer.
Deep Learning-Based Prediction and Classification of Microplastics in Water Samples
AI summary Read the abstract
Researchers built an AI tool that can automatically spot and identify tiny plastic particles in drinking water samples, correctly catching them about 92% of the time — much faster than manual lab methods. This matters because it could help water utilities quickly find contamination hotspots and monitor water quality in real time, making it easier to protect drinking water from microplastic pollution before it reaches your tap.
MicroNet: Automated Classification of Holographic Microplastics and Environmental Impact Analysis using Convolutional Neural Networks and Large Language Models
AI summary Read the abstract
Scientists built an AI tool that can quickly identify different types of microplastic pollution in water samples using special camera images, achieving over 93% accuracy—much faster than traditional lab methods that require expert analysis. This matters because microplastics are increasingly found in our oceans, food, and even our bodies, and faster detection tools could help researchers and environmental agencies track pollution sources and respond more quickly to protect ecosystems (and ultimately, the food chain we depend on).
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.