cotton_weed_detection
收藏资源简介:
Cotton Weed Detection 是一个专门用于棉花田杂草检测的目标检测数据集。该数据集包含262张通过无人机拍摄的高分辨率图像,共计50,812个边界框标注,覆盖两个主要类别:杂草和棉花。数据标注过程使用LabelImg和Roboflow工具完成,确保标注质量。该数据集适用于农业领域的杂草识别与目标检测任务,可用于训练和评估计算机视觉模型在棉花田环境中的杂草检测能力。数据集以CC-BY-4.0许可证发布,总大小约为2.19GB,包含训练集262个样本,旨在支持农业自动化和精准农业研究。
Cotton Weed Detection is a target detection dataset for weed detection in cotton fields. It contains 262 drone-captured images with a total of 50,812 bounding box annotations, covering two categories: weeds and cotton. The annotations were generated using LabelImg and Roboflow tools. This dataset is suitable for weed identification and object detection tasks in agriculture, and can be used to train and evaluate computer vision models for weed detection in cotton field environments. Released under the CC-BY-4.0 license, the dataset has a total size of approximately 2.19GB and includes 262 training samples.
数据集概述:棉花杂草检测(Cotton Weed Detection)
- 任务类型:目标检测(Object Detection)
- 许可协议:CC-BY-4.0
- 数据集规模:样本数少于1000(n<1K)
数据构成
- 图像总数:262张
- 标注框总数:50,812个
- 类别数量:2类(weed 和 cotton)
- 划分方式:仅包含训练集(train),共262个样本
数据特征
- image:图像数据(dtype: image)
- objects:包含以下字段
- bbox:边界框坐标(浮点型列表)
- categories:类别ID(整型列表)
- category_names:类别名称(字符串列表)
数据集大小
- 下载大小:约2.19 GB(2,192,509,526 bytes)
- 数据集总大小:约2.19 GB(2,192,609,502 bytes)
引用信息
- 论文:Das et al., 2026, "A UAV Image Dataset for Object Detection with Annotations Generated Using LabelImg and Roboflow", Data in Brief
- 数据源:Mendeley Data, doi: 10.17632/sx2tphzvcw.1
该数据集是 AgML 项目的一部分,索引于 https://project-agml.github.io/。



