MinneApple: A Benchmark Dataset for Apple Detection and Segmentation

We present a new dataset with the goal of advancing the state-of-the-art in fruit detection, segmentation, and counting in orchard environments. We hope to achieve this by providing a large variety of high-resolution images acquired in orchards, together with human annotations of the fruits on the trees. Objects are labeled using polygon masks for each object instance to aid in precise object detection, localization or segmentation. Additionally, we provide data for patch-based counting of clustered fruits. Our dataset contains over 40'000 annotated object instances in 1000 images.

本研究提出一款全新数据集，旨在提升果园场景下水果检测、实例分割与计数任务的现有最优性能。为实现该目标，我们提供了多组果园实地采集的高分辨率图像，并附带果树果实的人工标注信息。所有目标均采用多边形掩码（polygon masks）为每个目标实例完成标注，以支撑精准的目标检测、定位或分割任务。此外，本数据集还提供面向簇生果实的基于图像块的计数任务所需数据。本数据集共包含1000张图像，总计超过40000个带标注的目标实例。