云/天空分割

1. 简介该数据集专为人工智能研究提供支持，包含 500 多张带标签的天空和云朵相关图像，由 Acme AI 在孟加拉国夏季（热带与温带区交界处）收集，涵盖低云、中云、高云场景；图像按实例分割原则标注，包含 4 个标准类别（天空、云、太阳、鸟类），标注工具采用 SuperAnnotate 的像素编辑器，确保标注精准性1. Introduction: This dataset is developed to support artificial intelligence research. It contains over 500 labeled images related to skies and clouds, collected by Acme AI during the summer in Bangladesh, which is located at the junction of tropical and temperate zones. The dataset covers scenarios of low-level, mid-level and high-level clouds. All images are annotated in accordance with the instance segmentation principle, with four standard categories: sky, cloud, sun and bird. The pixel-level editor from SuperAnnotate was adopted as the annotation tool to ensure the accuracy of annotations.

搜集汇总

背景与挑战

背景概述

该数据集包含500多张带标签的天空和云朵图像，涵盖多种云场景，标注了4个标准类别，适用于无人机导航和天气预报模型的训练。