Beijing Laboratory of Food Quality and Safety, College of Engineering, China Agricultural University, Beijing, China.
College of Information and Electrical Engineering, China Agricultural University, Beijing, China.
J Sci Food Agric. 2024 Jan 15;104(1):273-285. doi: 10.1002/jsfa.12916. Epub 2023 Aug 21.
Consumers all throughout the world enjoy kiwifruit. After harvest, there are as much as 20-25% of kiwifruit lost along the entire industrial chain. An intelligent flexible manipulator system based on flexible tactile sensing (IFMSFTS) was created to automatically and intelligently classify kiwifruit ripeness in order to minimize loss.
The flexible manipulator is coupled with the flexible tactile sensor. When kiwifruits are being gripped by the manipulator, the flexible sensor perceives their firmness, and the mapping relationship between firmness and ripeness allows for the prediction and evaluation of the kiwifruit's ripeness. Principal component analysis (PCA) is employed to minimize the dimension of the sample firmness data set. K-Nearest neighbor (KNN) and support vector machine (SVM) classifiers are utilized to train and test the data. The findings indicate that PCA-KNN's classification accuracy is 97.5% and PCA-SVM's classification accuracy is 96.24%. The first is a better fit.
IFMSFTS can precisely classify ripeness, effectively address the issue of fruit loss, and realize the sustainable and clean production of fruit by sensing the firmness of kiwifruit and relying on the mapping link between firmness and ripeness. © 2023 Society of Chemical Industry.
世界各地的消费者都喜欢猕猴桃。收获后，整个产业链中猕猴桃的损失高达 20-25%。为了最大限度地减少损失，创建了一种基于柔性触觉传感的智能灵活操纵器系统（IFMSFTS），以实现猕猴桃成熟度的自动和智能分类。
柔性操纵器与柔性触觉传感器相耦合。当操纵器夹持猕猴桃时，柔性传感器感知其硬度，并且硬度与成熟度之间的映射关系允许对猕猴桃的成熟度进行预测和评估。主成分分析（PCA）用于最小化样本硬度数据集的维度。K-最近邻（KNN）和支持向量机（SVM）分类器用于训练和测试数据。结果表明，PCA-KNN 的分类准确率为 97.5%，PCA-SVM 的分类准确率为 96.24%。前者更合适。
IFMSFTS 可以精确地分类成熟度，通过感知猕猴桃的硬度并依赖硬度与成熟度之间的映射关系，有效解决水果损失问题，实现水果的可持续和清洁生产。© 2023 化学工业协会。
