论文部分内容阅读
基于美国USDA苹果分级标准研制了苹果擦伤检测分级系统样机。系统依据所研究的苹果表面擦伤拉曼光谱PLS/CVA处理和UNEQ模型分类,通过设计的LabVIEW虚拟仪器控制系统实时完成苹果分级。对48个预先分级的美国红元帅苹果(reddelicious#4167)分别进行两轮实际分级实验来评估该系统的性能。实验结果表明该分级检测系统对无擦伤、轻擦伤和重擦伤苹果组的分级平均准确率分别为100%、99%、97%。分级系统每分钟可分级苹果3～6个,对运动苹果分级误判率较低,是实施苹果自动检测分级生产线的基本模块。
Based on the United States USDA apple grading standards developed apple scratch test grading system prototype. Based on the PLS / CVA processing and UNEQ model classification of apple surface abrasion Raman spectra, the system realizes apple grading through the designed LabVIEW virtual instrument control system. Forty-eight pre-graded American red apple (# 4167) were subjected to two actual grading experiments to assess the performance of the system. The experimental results showed that the graded average accuracy of the grading system was 100%, 99% and 97% respectively for the apple group without scratches, bruises and scratches. The grading system can grade 3 to 6 apples per minute, and has a low false-positive rate of apples grading. It is the basic module for implementing Apple’s automated grading production line.