POLLEN73S: An image dataset for pollen grains classification

Authors

Gilberto Astolfi, Universidade Federal de Mato Grosso do Sul
Ariadne Barbosa Gonçalves, Universidade Federal de Mato Grosso do Sul
Geazy Vilharva Menezes, Universidade Federal de Mato Grosso do Sul
Felipe Silveira Brito Borges, Universidade Catolica Dom Bosco
Angelica Christina Melo Nunes Astolfi, Universidade Federal de Mato Grosso do Sul
Edson Takashi Matsubara, Universidade Federal de Mato Grosso do Sul
Marco Alvarez, University of Rhode IslandFollow
Hemerson Pistori, Universidade Federal de Mato Grosso do Sul

Document Type

Article

Date of Original Version

11-1-2020

Abstract

The classification of pollen species and types is an important task in many areas like forensic palynology, archaeological palynology, and melissopalynology. This paper presents a new public annotated image dataset for the Brazilian Savanna called POLLEN73S composed of 2523 images from 73 pollen types. Using the state-of-the-art Convolutional Neural Networks (CNNs), we provide a baseline for pollen grain classification. Our experiments showed evidence that DenseNet-201 and ResNet-50 have superior performance against the other CNNs tested, achieving precision results of 95.7% and 94.0%, respectively. Due to its category coverage and satisfactory diversity of examples, POLLEN73S offers a diversity of pollen grain to guide progress in computer vision to solve Palynology problems.

Publication Title, e.g., Journal

Ecological Informatics

Volume

60

Citation/Publisher Attribution

Astolfi, Gilberto, Ariadne B. Gonçalves, Geazy V. Menezes, Felipe S. Borges, Angelica C. Astolfi, Edson T. Matsubara, Marco Alvarez, and Hemerson Pistori. "POLLEN73S: An image dataset for pollen grains classification." Ecological Informatics 60, (2020). doi: 10.1016/j.ecoinf.2020.101165.