OliveTreeCrownsDb: A High-Resolution Dataset for Agricultural Computer Vision
Description
The OliveTreeCrown dataset contains high-resolution images divided into various grid configurations: 1×1 (original), 3×3, 6×6, and 9×9. Each segment is thoroughly annotated to ensure accurate object detection, providing precise and detailed labeling of olive tree crowns. In addition to the annotated image data, the dataset contains a point cloud representation, a Digital Elevation Model (DEM), and spatial data in Keyhole Markup Language (KML) format. These components collectively capture the three-dimensional geometry, topographic features, and geospatial characteristics of the study area. The XYZ coordinates in the point cloud data define the precise spatial position of each point, contributing to a comprehensive spatial representation. By integrating 3D data and geospatial attributes, this dataset offers a valuable resource for advanced spatial modeling and analysis. It serves as a solid foundation for applications such as multi-scale analysis, 3D mapping, and precision agriculture, fostering innovation in remote sensing and AI-driven agricultural solutions.
Files
Steps to reproduce
In the Meknes region of Morocco, a DJI Phantom 4 RTK drone captured 46 high-resolution images over the study farm at an altitude of 74.6 meters. To process and organize this data for analysis, follow these steps: 1.Data Export and Segmentation: Segment each image into grids of 1×1 (original), 3×3, 6×6, and 9×9 to facilitate multi-scale analysis. 2.Metadata Extraction: Retrieve key metadata such as date, time, GPS coordinates, altitude, and weather conditions from each image to enrich the dataset with essential contextual information. 3.Generating Point Cloud, DEM, and Orthophoto: Generate a dense point cloud to capture detailed terrain features. Create a Digital Elevation Model (DEM) to represent elevation and topography. Produce an orthophoto to provide an accurate, georeferenced image of the study area, useful for spatial analysis.
Institutions
- Universite Sidi Mohamed Ben Abdellah Faculte des Sciences Dhar El Mahraz-Fes