Phenotyping is essential in the process of varietal selection. In the case of sugar beets, richness (g/100g), that is, sugar content, is the key information. The need to acquire this information in a rapid, non-destructive and cheap manner leads the sugar industry to look for portable solutions that enable the suitable field measurements. In this work, a low-cost handheld and narrow visible-NIR spectral range microspectrometer is assessed for its ability to provide such information. During a two-year campaign from 2017 to 2018, 649 samples of sugar beet were measured. The resulting data, along with the reference values for richness, were used to build a predictive model with partial least squares (PLS) regression. Acceptable performance in the estimation of richness from both 2017 data (SEP = 0.84 g/100 g) and 2018 data (SEP = 0.90 g/100 g) is achieved. This study also shows that updating the spectral database is possible by calibration transfer models. From the different tested transfer strategies, the combination of model update and slope-bias correction achieves the best performance, demonstrating that the use of 2017 model on different years is possible and only 75 new sugar beets are necessary to guarantee a richness error lower than 1.05 g/100 g. This work suggests that the molecular sensor could offer a useful tool for a rapid, low cost and non-destructive prediction of richness in sugar beets.

在品种选择过程中，表型分析是必不可少的。就甜菜而言，丰富度（g/100g），即糖含量，是关键信息。以快速、非破坏性和廉价的方式获取这些信息的需求促使制糖业寻找能够进行适当现场测量的便携式解决方案。在这项工作中，评估了一种低成本的手持式窄可见近红外光谱范围显微光谱仪提供此类信息的能力。在 2017 年至 2018 年的两年活动中，共测量了 649 个甜菜样品。所得数据以及丰富度的参考值用于构建具有偏最小二乘 (PLS) 回归的预测模型。从 2017 年数据 (SEP = 0.84 g/100 g) 和 2018 年数据 (SEP = 0. 达到 90 g/100 g)。该研究还表明，通过校准传递模型更新光谱数据库是可能的。从不同的测试转移策略来看，模型更新和斜率偏差校正的组合达到了最佳性能，表明在不同年份使用 2017 模型是可能的，只需 75 个新甜菜即可保证低于 1.05 的丰富度误差克/100 克。这项工作表明，分子传感器可以提供一种有用的工具，用于快速、低成本和无损地预测甜菜的丰富度。证明在不同年份使用 2017 年模型是可能的，只需 75 个新甜菜即可保证丰度误差低于 1.05 g/100 g。这项工作表明，分子传感器可以提供一种有用的工具，用于快速、低成本和无损地预测甜菜的丰富度。证明在不同年份使用 2017 年模型是可能的，只需 75 个新甜菜即可保证丰度误差低于 1.05 g/100 g。这项工作表明，分子传感器可以提供一种有用的工具，用于快速、低成本和无损地预测甜菜的丰富度。