Benchmarking new methods for estimation of weight of mango fruit on tree

This research aimed to recommend methods, equipment, and protocols to Australian mango farmers for the forward estimation of at-harvest fruit load, with a focus on fruit size forecasting. A comparison of field temperature measurement systems for growing degree days (GDD) estimation was undertaken, with recommendations based on accuracy and cost. Existing GDD targets for key cultivars were validated. Recommendations for allometric models to estimate fruit weight from dimensions were made, with cultivar-specific parameters. A comparison of fruit growth models was undertaken, with recommendations for forecasting accuracy at different pre-harvest intervals. Recommendations for improvements to machine vision-based fruit sizing tools were also made. Together, these activities provided a holistic approach to forecasting fruit size at harvest.本研究旨在为澳大利亚芒果种植户提供可用于收获期果实负载量前瞻性估算的方法、设备与操作规范，重点聚焦果实大小预测。本研究针对用于生长度日（GDD）估算的田间温度测量系统开展对比分析，并基于精度与成本维度提出推荐方案。对核心芒果品种的现有生长度日阈值进行了有效性验证。提出了基于维度参数估算果实重量的异速生长模型推荐方案，并附带各品种专属参数。开展了果实生长模型的对比研究，并针对不同收获前时间区间的预测精度给出适配建议。同时针对基于机器视觉的果实尺寸检测工具提出了优化方案。上述各项工作共同构建了一套完整的收获期果实大小预测体系。