Saved in:
Bibliographic Details
Main Authors: Staforelli-Vivanco, J., Jofré, R., Muñoz, B., Salamanca, V., Coelho, P., Sanhueza, I., Viafora, L., Toro, C., Troncoso, J., Rondanelli-Reyes, M., Lamas, I.
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.16743
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908975735242752
author Staforelli-Vivanco, J.
Jofré, R.
Muñoz, B.
Salamanca, V.
Coelho, P.
Sanhueza, I.
Viafora, L.
Toro, C.
Troncoso, J.
Rondanelli-Reyes, M.
Lamas, I.
author_facet Staforelli-Vivanco, J.
Jofré, R.
Muñoz, B.
Salamanca, V.
Coelho, P.
Sanhueza, I.
Viafora, L.
Toro, C.
Troncoso, J.
Rondanelli-Reyes, M.
Lamas, I.
contents Traditional melissopalynology is a time-consuming and subjective process, often taking 4-6 hours per sample. We present an automated, high-throughput microscopy system that integrates $H\infty$ robust mechanical control with advanced deep learning pipelines for the precise counting, classification, and morphological analysis of pollen grains from Bio Bio region in south central territory in Chile. Our system employs $U^{2}$-Net for salient object detection and a DINOv2 Vision Transformer backbone trained via Deep Metric Learning for classification. By integrating Gradient-Weighted Attention, the model provides human-interpretable texture and diagnostic feature annotations. The system achieves a 95.8$\%$ classification recall and a 6x processing speedup compared to manual expert analysis.
format Preprint
id arxiv_https___arxiv_org_abs_2604_16743
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Automated Palynological Analysis System: Integrating Deep Metric Learning and $U^{2}$-Net Detection in $H\infty$ bright field microscopy
Staforelli-Vivanco, J.
Jofré, R.
Muñoz, B.
Salamanca, V.
Coelho, P.
Sanhueza, I.
Viafora, L.
Toro, C.
Troncoso, J.
Rondanelli-Reyes, M.
Lamas, I.
Computer Vision and Pattern Recognition
Optics
Traditional melissopalynology is a time-consuming and subjective process, often taking 4-6 hours per sample. We present an automated, high-throughput microscopy system that integrates $H\infty$ robust mechanical control with advanced deep learning pipelines for the precise counting, classification, and morphological analysis of pollen grains from Bio Bio region in south central territory in Chile. Our system employs $U^{2}$-Net for salient object detection and a DINOv2 Vision Transformer backbone trained via Deep Metric Learning for classification. By integrating Gradient-Weighted Attention, the model provides human-interpretable texture and diagnostic feature annotations. The system achieves a 95.8$\%$ classification recall and a 6x processing speedup compared to manual expert analysis.
title Automated Palynological Analysis System: Integrating Deep Metric Learning and $U^{2}$-Net Detection in $H\infty$ bright field microscopy
topic Computer Vision and Pattern Recognition
Optics
url https://arxiv.org/abs/2604.16743