Stage-Wise Field Observation Framework for Early Mango Yield Estimation

Mango (Mangifera indica) is one of the most important fruit crops cultivated in tropical and subtropical regions. Accurate yield estimation is essential for harvest planning, supply-chain management, and orchard decision-making. Conventional yield estimation practices primarily rely on manual observations, which are often labor-intensive and subject to observer variability. Recent advances in computer vision and remote sensing have created opportunities for data-driven yield forecasting; however, the availability of structured field datasets representing flowering and early fruit development stages remains limited. This study presents a stage-wise field observation framework for early mango yield estimation based on systematic image acquisition across multiple phenological stages. A longitudinal dataset was developed using 25 mango trees monitored from four viewing directions at weekly intervals. The framework incorporates image acquisition, quality verification, preprocessing, stage annotation, and dataset organization procedures. The resulting dataset captures flowering progression and early fruit development under real orchard conditions. The framework establishes a foundation for future yield estimation studies and contributes a structured methodology for collecting temporal phenological information in mango orchards. The proposed approach may support future research in crop monitoring, phenological assessment, and early-season yield forecasting.