高通量水果快速无损检测技术与智能分选系统

浙江大学与合作单位开发 AI+ 可见 / 近红外成像融合的水果外观品质机器视觉快速检测、基于连续积分的小空间全透射水果内部品质可见 / 近红外光谱无损检测、动态参比 -AI 辨识 -特定波段信噪比增强的水果光谱弱信号稳定获取、面向不同属性因子差异的基于迁移学习策略的模型更新与传递、多种包装形式 - 多通道合流 - 大通量综合分选的智能分选策略等，创制了系列水果品质智能检测分选装备，实现水果内外部品质检测分选的防损伤、测得全、测得准、

高速度、智能化和少人化，可用于水果内外部多品质指标的同步无损检测与分选，为农产品增值减损和农业增效提供技术与装备支撑。
 

Zhejiang University, in collaboration with its partners, has invented machine vision-based rapid detection of fruit external quality utilizing AI + visible/near-infrared imaging fusion, non-destructive assessment of fruit internal quality using full-transmission visible/near-infrared spectroscopy with continuous integration in confined spaces, stable acquisition of weak spectral signals in fruits through dynamic reference calibration, AI recognition, and signal-to-noise ratio enhancement at specific wavelengths, model updating and transfer based on learning strategies to address variations in quality attributes, and intelligent sorting strategies incorporating multiple packaging formats, multi-channel convergence, and high-throughput integrated sorting. These technologies enable damage-free, comprehensive, high-precision, high-speed, intelligent, and automated detection and sorting of both internal and external fruit qualities. The system facilitates simultaneous nondestructive detection and sorting of multiple fruit quality parameters, providing technological and equipment support for value addition, loss reduction, and efficiency improvement in agriculture.

水果内外部品质，无损检测，高通量，成套智能装备

Fruit Internal and External Quality; Nondestructive Detection; High Throughput; Intelligent Grading Equipment

完成单位

浙江大学

Zhejiang University