利用深度学习进行实时高光谱成像田间草莓成熟度评估

摘要：草莓是一种营养丰富的流行水果。本研究使用高光谱成像 (HSI) 系统在田间和实验室条件下评估了这种水果的成 熟度。收集了早熟和成熟期草莓的 HSI 数据，覆盖波长范围为 370 至 1015nm。使用顺序特征选择 (SFS) 算法选择光 谱特征波长。分别为田间 (530 和 604nm) 和实验室 (528 和 715nm) 样品选择了两个波长。然后，基于支持向量机 (SVM) 分类器验证了此类光谱特征的可靠性。SVM 分类模型的性能良好，田间和实验室条件下样品的接收机操作特征 值均高于 0.95。同时，针对实验室样品，从光谱特征波长和前三个主成分中提取空间特征图像。使用预训练的 AlexNet 卷积神经网络（CNN）对早熟和成熟草莓样品进行分类，对测试数据集的准确率达到 98.6%。上述结果表明，实时 HSI 系统有望在田间和实验室条件下评估草莓的成熟度，这可能是评估农民和生产者收获时间管理的潜在应用技术。

关键词：草莓成熟度 高光谱影像 田间CNN

期刊：Artificial Intelligence in Agriculture

DOI：https://DOI.org/10.1016/j.aiia.2020.04.003

设备：GaiaField

Real-time hyperspectral imagingforthein-field estimation of strawberry ripeness with deep learning

Abstract：Strawberry is one of the popular fruits with numerous nutrients. The ripeness of this fruits was estimated using the hyperspectral imaging (HSI) system in field and laboratory conditions in this study. Strawberry at early ripe andripestageswerecollectedHSIdata,coveredwavelengthrangesfrom370to1015nm.Spectralfeaturewave lengths were selected using the sequential feature selection (SFS) algorithm. Two wavelengths selected for field (530 and 604 nm)andlaboratory (528 and 715nm) samples, respectively. Then, reliability of such spectral fea tures wasvalidated basedonsupportvectormachine(SVM)classifier. PerformanceofSVMclassificationmodels hadgoodresults withreceiver operating characteristic values for samples underbothfield and laboratory condi tions higher than 0.95. Meanwhile, the spatial feature images were extracted from the spectral feature wave length and the first three principal components for laboratory samples. Pretrained AlexNet convolutional neural network(CNN)wasusedtoclassify theearly ripe and ripe strawberry samples, which obtained the accu racy of 98.6% for test dataset. The above results indicated real-time HSI system was promising for estimating strawberry ripeness under field and laboratory conditions, which could be a potential application technique for evaluating the harvesting time management for farmers and producers.

Keywords：Strawberry ripeness Hyperspectral imagery In field CNN