Orchard Environments

This dataset has been curated to support the development and evaluation of lightweight object detection algorithms for precision pesticide spraying in orchard environments. It comprises annotated images captured in diverse natural lighting and occlusion conditions typical of real-world agricultural fields. The dataset includes high-resolution RGB images along with corresponding bounding box annotations in YOLO format, identifying key targets. All annotations are manually verified to ensure label accuracy. The dataset is formatted to be compatible with popular deep learning frameworks and includes configuration files for seamless training and testing. By enabling robust training and benchmarking of object detection models in unstructured environments, this dataset contributes to the advancement of intelligent agricultural robotics and vision-based automation.