Temporal Analysis of Multi-Stage Flowering Progression for Mango Yield Forecasting

Mango yield estimation plays an important role in orchard management and agricultural planning. Conventional approaches often depend on manual field observations, which are time-consuming and may vary among observers. Recent studies have demonstrated the usefulness of computer vision, remote sensing, and data-driven techniques for agricultural monitoring. However, limited attention has been given to understanding how flowering progression over multiple growth stages can contribute to yield forecasting. This study investigates the temporal characteristics of mango flowering progression using sequential field observations collected from orchard environments. A structured image acquisition strategy was adopted, involving 25 trees monitored from four viewing directions across multiple developmental stages. The collected images were processed by resizing, region-of-interest extraction, colour normalisation, and stage encoding. The study establishes a framework for analyzing flowering-to-fruit transitions and examines the feasibility of using temporal phenological information for yield forecasting. Observations indicate that flowering progression exhibits measurable visual changes that may serve as useful indicators for future yield estimation studies. The proposed framework contributes a systematic methodology for longitudinal orchard monitoring and provides a foundation for future predictive modeling efforts.