A Public Image Database for Benchmark of Plant Seedling Classification Algorithms

TL;DR

This paper introduces a publicly available image database of 960 plant seedlings across 12 species, along with a benchmark based on F1 scores to standardize classification evaluation.

Contribution

It provides a new, annotated plant seedling image dataset and a benchmark framework for evaluating classification algorithms.

Findings

Dataset contains 960 images of 12 species

Benchmark based on F1 scores established for classification

Dataset and benchmark available online

Abstract

A database of images of approximately 960 unique plants belonging to 12 species at several growth stages is made publicly available. It comprises annotated RGB images with a physical resolution of roughly 10 pixels per mm. To standardise the evaluation of classification results obtained with the database, a benchmark based on $f_{1}$ scores is proposed. The dataset is available at https://vision.eng.au.dk/plant-seedlings-dataset