0
Article ? AI-assigned paper type based on the abstract. Classification may not be perfect — flag errors using the feedback button. Tier 2 ? Original research — experimental, observational, or case-control study. Direct primary evidence. Environmental Sources Sign in to save

Encoding Holographic Data Into Synthetic Video Streams for Enhanced Microplastic Detection

IEEE Access 2025 Score: 38 ? 0–100 AI score estimating relevance to the microplastics field. Papers below 30 are filtered from public browse.
Paolo Russo, Fabiana Di Ciaccio, Pasquale Santaniello, Teresa Cacace, Pierluigi Carcagnì, Marco Del Coco, Melania Paturzo

Summary

Researchers developed a deep learning pipeline for microplastic detection using digital holography, introducing an enhanced dataset (HMPD 2.0) with 24 amplitude and phase channels and a pseudo-RGB compression technique to efficiently detect and classify synthetic microfibers released during laundry washing.

This study presents a novel deep learning pipeline for the detection and classification of microplastics using digital holography, with a focus on synthetic microfibers released during laundry. We introduce HMPD 2.0, an enhanced version of the Holography MicroPlastic Dataset featuring 24 amplitude and phase channels per sample, obtained through varying spatial filtering, numerical aberration correction, and propagation. To reduce input dimensionality, we propose a pseudo-RGB compression technique that groups grayscale channels into synthetic frames, which are then interpreted as a video sequence. This allows the use of transformer-based video architectures, particularly TimeSformer, for spatiotemporal modeling. Experimental results demonstrate that TimeSformer achieves a classification accuracy of up to 97.91%, with compressed 8-frame inputs maintaining high performance while significantly reducing inference time (from 42 ms to 16 ms per sample). These findings validate the effectiveness and efficiency of our approach, which supports real-time deployment on edge devices.

Sign in to start a discussion.

More Papers Like This

Article Tier 2

Microplastic pollution monitoring with holographic classification and deep learning

This study used digital holographic microscopy combined with deep learning to classify microplastic particles in water samples, achieving high classification accuracy and demonstrating the potential for automated, high-throughput microplastic monitoring.

Article Tier 2

Automatic Detection of Microplastics by Deep Learning Enabled Digital Holography

Researchers developed a digital holography system combined with deep learning to automatically detect and identify microplastics in water without manual image analysis. The system processes raw holographic images directly, offering a faster and more scalable approach to microplastic monitoring in environmental samples.

Article Tier 2

Micro-Objects Classification for Microplastic Pollution Detection using Holographic Images

Researchers developed a machine learning system that uses holographic 3D images to automatically classify microplastics in water samples, distinguishing them from other microscopic particles with high precision. Current microplastic monitoring is slow and labor-intensive, so automated detection tools are essential for large-scale environmental surveillance. This approach could significantly speed up the monitoring of microplastic pollution in aquatic environments.

Article Tier 2

Deep Classification of Microplastics Through Image Fusion Techniques

Deep neural networks were applied to classify microplastic fibers captured via digital holography microscopy, using image fusion techniques on the Holography Micro-Plastic Dataset benchmark. The study demonstrated promising accuracy for distinguishing microplastics from other debris, advancing automated microplastic identification in water quality monitoring.

Article Tier 2

Digital holographic microplastics detection and characterization in heterogeneous samples via deep learning

Researchers used digital holographic microscopy combined with deep learning to detect and characterize microplastic particles in heterogeneous samples containing algae, microorganisms, and other natural particles. This automated approach could improve the speed and accuracy of environmental microplastic monitoring.

Share this paper