Avo-AirDB: An avocado UAV Database for agricultural image segmentation and classification

Abstract

Unmanned aerial vehicles (UAVs) with on-board cameras have the advantage of providing Bird-view images (Aerial images). This type of image is considered as a rich source of information especially for intelligent agriculture. A dataset of 984 aerial images of avocado threes is made publicly available with a ground resolution of 2.7 cm per pixel. It has been collected from over 113 Hectares of an avocado farm in ALLAL TAZI region of Morocco using a DJI Phantom 4 Pro UAV. It comprises original bird view and annotated images. The dataset is available at https://data.mendeley.com/datasets/tvhh83r3hj/2

Keywords

• Dataset
• Avocado
• Unmanned aerial vehicles
• Digital agriculture