食品与机械2025，Vol.41Issue(12)：59-65,7.DOI:10.13652/j.spjx.1003.5788.2024.60160
多信息集成技术在脐橙糖度无损检测中的应用
Application of multi-information integration technology in nondestructive detection of navel orange sugar content
摘要
Abstract
[Objective]To solve the problem of low detection accuracy in existing nondestructive detection methods for navel orange sugar content.[Methods]Based on the analysis of detection schemes,a multi-information integration method for nondestructive detection of navel orange sugar content is proposed.Data collection is conducted using spectral detection technology,machine vision technology,and electronic nose technology.Spectral data is obtained by competitive adaptive weighted sampling of 17 wavelength variables.The principal component analysis is applied to extract 6 features from machine vision data and 4 features from electronic nose sensor data,which are then used as inputs to an improved RBF neural network model for sugar content detection.[Results]Compared with conventional detection methods,the proposed multi-information integration method extracts features more comprehensively,resulting in higher detection accuracy and efficiency.The coefficient of determination is 0.960 8,the root mean square error is 0.083 2 °Brix,and the average detection time is 0.154 s.[Conclusion]This scheme improves the detection accuracy of navel orange sugar content and has certain reference value.关键词
无损检测/脐橙/糖度/光谱检测技术/机器视觉技术/电子鼻技术Key words
nondestructive detection/navel orange/sugar content/spectral detection technology/machine vision technology/electronic nose technology引用本文复制引用
何晓田,王文凡,申杰,田二林..多信息集成技术在脐橙糖度无损检测中的应用[J].食品与机械,2025,41(12):59-65,7.基金项目
河南省科技攻关项目(编号:232102211072) （编号:232102211072）
