unileon-robotics/Veridis
收藏Hugging Face2025-11-21 更新2026-01-03 收录
下载链接:
https://hf-mirror.com/datasets/unileon-robotics/Veridis
下载链接
链接失效反馈官方服务:
资源简介:
---
license: cc-by-sa-4.0
task_categories:
- robotics
- object-detection
tags:
- robotics
- computer-vision
- object-detection
- agriculture
---
# VERIDIS Dataset
## Overview
This repository contains the **VERIDIS** dataset, a collection of annotated agricultural images for crop detection and identification. The dataset comprises field images of beet and corn crops captured by a ground-level robotic platform, organized in YOLO format for object detection tasks.
The dataset includes 10,077 total files distributed across training (9,100 files), validation (488 files), and test (489 files) sets, with images of two crop types: beet (Class 0) and corn (Class 1).
## Repository Contents
### Main Dataset
- **veridis.zip**
- Main dataset containing the complete collection of data for identity verification research
- Recently updated with the latest version of the dataset
- Format: Compressed ZIP archive
### Processing Scripts
#### 1. `anonymize_persons.py`
Script designed for anonymizing personal data within the dataset. It implements privacy-preserving techniques to protect personally identifiable information while maintaining data utility for research purposes.
#### 2. `data_augmentation.py`
Data augmentation script that expands the dataset through systematic transformations and variations generation.
#### 3. `extract_frames.py`
Tool for extracting individual frames from video sequences or temporal data.
#### 4. `train.py`
Main training script for machine learning and deep learning models on the VERIDIS dataset.
## License
This project is licensed under **CC-BY-SA-4.0** (Creative Commons Attribution-ShareAlike 4.0 International).
---
许可证:cc-by-sa-4.0
任务类别:
- 机器人学
- 目标检测(object detection)
tags:
- 机器人学
- 计算机视觉(computer vision)
- 目标检测
- 农业
---
# VERIDIS数据集
## 概述
本仓库包含**VERIDIS**数据集,这是一个用于作物检测与识别的带注释农业图像集合。该数据集包含由地面机器人平台拍摄的甜菜和玉米作物田间图像,以YOLO格式组织,适用于目标检测任务。
数据集共包含10,077个文件,分布于训练集(9,100个文件)、验证集(488个文件)和测试集(489个文件),涵盖两种作物类型:甜菜(类别0)和玉米(类别1)。
## 仓库内容
### 主数据集
- **veridis.zip**
- 包含身份验证研究完整数据集的主数据集
- 最近已更新至数据集的最新版本
- 格式:压缩ZIP归档
### 处理脚本
#### 1. `anonymize_persons.py`
用于匿名化数据集中个人数据的脚本。它采用隐私保护技术,在保护个人可识别信息的同时,保留数据的研究实用性。
#### 2. `data_augmentation.py`
通过系统性变换和变异生成来扩展数据集的数据增强脚本。
#### 3. `extract_frames.py`
用于从视频序列或时间数据中提取单帧的工具。
#### 4. `train.py`
用于在VERIDIS数据集上训练机器学习和深度学习模型的主训练脚本。
## 许可证
本项目采用**CC-BY-SA-4.0**(知识共享署名-相同方式共享4.0国际许可协议)授权。
提供机构:
unileon-robotics



