Updated the README with dataset details to improve discoverability on Hugging Face

README.md

@@ -1,9 +1,59 @@
 ---
 license: cc-by-4.0
+task_categories:
+- feature-extraction
+language:
+- en
+- de
+tags:
+- image
+- LiDAR
+- RGB
+- Depth
+- RGB-D
+- Leaf-Area
+- computer-vision
+- Cross-modal
+pretty_name: irriwell
+size_categories:
+- 100K<n<1M
 ---
-This dataset captures 3D point clouds of appple trees over the period of three vegetation seasons, recorded in Potsdam, Germany. 
-The 3D point clouds of the trees were measured at two wavelengths (660 nm, 905 nm) using light detection and ranging laser scanners. 
-In parallel to the recording of 3D point clouds, RGB-D sensor was employed. Satellite data were extracted for the orchard over the three years.
-Reference analysis contains manually scanned leaf area defoliating entire trees. 
-The file HuggingFace_Irriwell_RGBD_LiDAR.xlsx provides the overview on all files. 
-A data paper is under development, with initial statistics on the data recorded.
+
+# IRRIWELL
+**Irriwell** is a multimodal dataset for leaf area estimation and canopy characterization of apple trees under real orchard conditions. The dataset combines RGB-D imagery, LiDAR point clouds, weather observations, and Sentinel-2 satellite imagery, enabling research on multi-scale sensing, sensor fusion, plant phenotyping, and digital twins.
+
+Data were collected over three growing seasons (2021–2023) in an experimental apple orchard in Potsdam-Marquardt, Germany, covering multiple developmental stages of the trees. The dataset includes sparse reference leaf area measurements obtained through destructive sampling and manual leaf area analysis, providing ground truth for validating leaf area estimation methods.
+
+## Sensors and Data Acquisition
+Data were acquired using a mobile sensing platform equipped with:
+- Sick LMS-511/LRS4000 LiDAR (905 nm)
+- Pepperl+Fuchs LiDAR (660 nm)
+- Intel RealSense D435 RGB-D camera
+
+Each tree was observed from both sides of the row. For every observation, one LiDAR scan and 20 synchronized RGB-D views were acquired per side. In addition to ground-based sensing, weather records and Sentinel-2 satellite imagery were collected throughout the study period.
+
+## Apple Cultivars
+| Tree type | Apple Cultivar |
+|----------------------|----------------|
+| Slender Spindle      | Jonaprince     |
+| Slender Spindle      | Gala           |
+| Slender Spindle      | Idared         |
+| Bi-tree              | Gala           |
+| Bi-tree              | Red Sentinel   |
+| Pollinator           | Red Idared     |
+
+## Dataset Contents
+The dataset contains:
+- RGB images
+- Depth images
+- LiDAR point clouds (660 nm and 905 nm)
+- Weather data
+- Sentinel-2 satellite imagery
+- Reference leaf area measurements
+- Metadata and acquisition records
+
+The file `HuggingFace_Irriwell_RGBD_LiDAR.xlsx` provides an overview of all recordings and associated metadata.
+
+## Applications
+Potential use cases include:
+- Leaf area estimation (RGB, depth and LiDAR)