Work

Synthetic Apple Collection (SyntAC): render-based synthetic datasets for apple detection

Year: 2026

Type: data-paper

Abstract: We present the Synthetic Apple Collection (SyntAC), a 20 GB set of synthetic images developed for apple detection in RGB images. The collection consists of 3 datasets, each containing 3840 image samples, for a total of 11,520 images, generated using Unreal Engine 5 and annotated in the YOLO format. The datasets include a variety of scenes in a virtual environment representing an apple orchard in different illumination conditions. Annotations are produced procedurally from engine data, providing perfect, pixel-to-pixel labeling and ensuring full reproducibility of the annotation process. The collection is designed to be easily reusable for the development, testing, and benchmarking of object detection algorithms and, more generally, machine learning models for precision agriculture.

Source: Data in Brief

DOI: 10.1016/j.dib.2026.113368

Authors: Jhonny Hueller, Massimo Vecchio, Fabio Antonelli

Institutions: University of Trento, Fondazione Bruno Kessler

Language: English

Topic: Smart Agriculture and AI

Subfield: Plant Science

Field: Agricultural and Biological Sciences

Domain: Life Sciences

SDG: Zero hunger

Open Access status: gold

Estimated APC paid: $1,220