# Aerial Multispectral Drone Surveys of a Sunflower Crop's Lifecycle (2023)

## Facts

- Provider: Sofia University Faculty of Mathematics and Informatics; SAP Labs Bulgaria
- License: CC-BY-4.0 (https://huggingface.co/datasets/su-fmi/msi-drone-crop-surveys)
- Size: 100,000 files
- Format: JPEG, TIFF (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/su-fmi/msi-drone-crop-surveys
- Access: public, readable without an account
- Category: Earth Observation & Geospatial
- Source: https://huggingface.co/datasets/su-fmi/msi-drone-crop-surveys
- Page: https://scigantic.com/archive/3a35e415-eec3-597a-8b96-01b41bab5b41
- Keywords: Earth Observation & Geospatial, satellite imagery, time series, JPEG, TIFF, Hugging Face

## About

Repeated multispectral drone surveys tracking one sunflower field through its whole lifecycle, from 25 April to 4 September 2023, produced by Sofia University's Faculty of Mathematics and Informatics with SAP Labs Bulgaria for research on precision agriculture. A DJI P4 multispectral drone captured imagery in visible and near-infrared bands, and composites were generated with DJI Terra at 70% frontal and 60% side overlap. Each survey is a folder named by date (DD.MM.YYYY; the root lists more than 40 survey dates, measured 2026-11-14). Within it, raw/ holds georeferenced composite images per band (result_Blue, result_Green and so on) with .prj projection and .tfw world files, and results/ holds per-flight subfolders where each sensing point on the mission path has one JPG and five grayscale TIF images, one per sensor. The repository uses about 457 GB. Uses include vegetation index time series (NDVI, NDRE), crop growth and stress monitoring, drone and Sentinel-2 cross-calibration and machine learning on multispectral imagery. Read the GeoTIFF-style rasters with rasterio or GDAL and the images with PIL. The card states access and use constraints as CC BY 4.0.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="su-fmi/msi-drone-crop-surveys", filename="01.08.2023/raw/100FPLAN/DJI_0010.JPG", repo_type="dataset")
print(path)
```

## Data structure

File types in sampled files: JPEG, TIFF

Example file: `01.08.2023/raw/100FPLAN/DJI_0010.JPG`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/3a35e415-eec3-597a-8b96-01b41bab5b41
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/3a35e415-eec3-597a-8b96-01b41bab5b41
- Over MCP: `get_schema_card` with id `3a35e415-eec3-597a-8b96-01b41bab5b41` at https://api.scigantic.com/mcp

## Related topics

- [Earth Observation & Geospatial datasets](https://scigantic.com/topics/earth-observation-and-geospatial)
- [Satellite imagery datasets](https://scigantic.com/topics/satellite-imagery)
- [TIFF datasets](https://scigantic.com/topics/tiff)
- [Time series datasets](https://scigantic.com/topics/time-series)
- [JPEG datasets](https://scigantic.com/topics/jpeg)
