A novel vision system with unique algorithms to recognise, count & size apples on trees to greatly improve crop forecasting and management to maximise yield and optimise market price - Applecount -

一种新颖的视觉系统，具有独特的识别、计数算法

• 批准号: 710522
• 金额: $ 12.14万
• 依托单位: THE TECHNOLOGY RESEARCH CENTRE LIMITED
• 依托单位国家: 英国
• 项目类别: GRD Proof of Concept
• 财政年份: 2014
• 资助国家: 英国
• 起止时间: 2014 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=710522
• 关键词: novel; vision; system; unique; algorithms

The UK is not self sufficient in apples, even during the high season, providing only one thirdof our own consumption, with the shortfall made up by imports. A large proportion of this isdue to our inability to meet demand for class1 fruit – the stringent specification set bysupermarkets, representing 80% of sales. Our client records show that similar orchards canhave outputs that vary by over 60% on the same cultivars, both on the overall tonnage yieldper hectare, and on the percentage of substandard fruit, with reject fruit going to waste or lowvalue processing and losing up to 80% of its value. A significant proportion of this variation isdown to management practice and crop forecasting.The market price is dependent on crop quality and the crop yield declared by the growers andwholesalers, and these estimates tend to be inaccurate with a variation of typically +/- 20%from final yield. This has a major impact on market price. Quotas are agreed with thesupermarkets early in the season, and must be fulfilled. Over-estimating yield meanspurchasing imports (at late in season high prices) to cover the shortfall, while underestimatingyield, means losing profits by selling excess crop to low value outlets or even forpigfeed. A key part of management practice is to know when and how to ‘thin’ crops topromote selective growth, and this is very dependent on knowing accurately how many applesthere are on each tree at various times in the growing season.Our client who represents one third of UK growers, believes that by standardising bestpracticeorchard management, and with a strategic approach to helping breed new cultivars,we could enable UK orchards to take back at least 100,000T of lost import volume, worth£50M.This project aims to create & prove the effectiveness of a novel vision based cropmeasurement technology for apple growers, capable of measuring apples while on the trees.

英国的苹果不能自给自足，即使在旺季也是如此，只提供了我们自己消费的三分之一，缺口由进口弥补。这在很大程度上是因为我们无法满足对1类水果的需求--这是超市设定的严格规格，占销售额的80%。我们的客户记录显示，类似的果园在相同品种上的产量差异可能超过60%，无论是在每公顷的总吨位产量上，还是在不合格水果的百分比上，都可能导致不合格水果浪费或低价值加工，并损失高达80%的价值。这种差异很大程度上取决于管理实践和作物预测。市场价格取决于作物质量和种植者和批发商宣布的作物产量，这些估计往往不准确，与最终产量的偏差通常为+/-20%。这对市场价格有重大影响。配额是在赛季初与超市达成一致的，必须完成。高估产量意味着高估进口(在季末高价时)以弥补缺口，而低估产量意味着将多余的作物出售给低价值的销售点甚至用作猪饲料而损失利润。管理实践的一个关键部分是知道何时以及如何‘瘦’庄稼以促进选择性生长，这非常依赖于准确地知道在生长季节的不同时间每棵树上有多少苹果。我们的客户代表了三分之一的英国种植者，他相信通过标准化最好的种植实践或果园管理，以及帮助培育新品种的战略方法，我们可以使英国果园收回至少10万吨损失的进口量，价值50 GB。该项目旨在为苹果种植者创造和证明一种新的基于视觉的作物测量技术的有效性，能够在树上测量苹果。