MangoSense: A time-series vision sensing dataset for mango tree segmentation and detection toward yield prediction

Open Webpage

Janaksinh Ven, Charu Sharma, Syed Azeemuddin

Published: 01 Jan 2025, Last Modified: 09 Nov 2025Comput. Electron. Agric. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Time-series mango orchard image dataset collected across multiple orientations.•Semi-automatic annotation using YOLO and SAM significantly reduces manual effort.•SegFormer and YOLO identified as top models for flower and fruitlet feature extraction.•Wind direction shown to influence flower-to-fruit conversion and yield variation.•Ground-truth fruit counts included to enable future yield prediction model development.