A High-Resolution Multifocal RGB Pollen Grain Image Dataset for Deep Learning Computer Vision Tasks from Biobío Region, Chile.
Sanhueza I, Coelho P, Viafora L, Jofre-Cerda R, Salamanca-Levi V
Pollinators
Native Chilean trees like the ulmo and quillay — whose pollen this dataset captures in unprecedented detail — are keystones of a temperate rainforest that holds more endemic species per acre than almost anywhere in South America, and better pollen ID tools mean faster early warnings when those plants start disappearing.
Scientists photographed pollen grains from 16 plant species found in southern Chile, capturing each grain at three different focus depths to reveal surface textures and internal structures that a single photo would miss. They used these images to teach a computer to recognize and outline each pollen grain with remarkable accuracy. The resulting photo library is now freely available so other researchers — and eventually apps — can automatically identify pollen from bees, air samples, or honey.
Key Findings
The dataset contains 16,198 high-resolution microscopy images (3088×2064 pixels) with 36,383 hand-verified pollen outlines across 16 species, making it one of the most detailed pollen datasets available.
Each pollen grain was photographed at three focal depths, capturing surface and interior features invisible in single-plane images — a design that improved AI classification robustness.
A baseline AI model trained on the dataset achieved 0.985 mask mAP@50 (near-perfect detection accuracy) on the validation set after just 50 training cycles.
chevron_right Technical Summary
Researchers in Chile built a large, high-quality image library of 16 pollen types from the Biobío Region, complete with precise shape outlines verified by a pollen expert. This dataset is designed to train AI systems that can automatically identify pollen — a task that matters for tracking plant diversity, allergies, and honey authenticity.
Abstract Preview
PollenBB16 is an RGB pollen image dataset of Chilean flora with pixel-accurate instance segmentation masks, whose annotation was fully verified by an expert palynologist to guarantee the taxonomic ...
open_in_new Read full abstractAbstract copyright held by the original publisher.
Species Mentioned
Was this useful?
Rapid disruption of pollination function by the invasive plant Impatiens glan...
That pretty pink-flowered plant spreading along your local riverbank or woodland edge could be quietly starving native wildflowers of the bee visits they nee...
The Valar are characters in J. R. R. Tolkien's Middle-earth writings. They are "angelic powers" or "gods" subordinate to the one God. The Ainulindalë describes how some of the Ainur choose to enter the world (Arda) to complete its material development, after its form is determined by the Music...