基于特征荧光信号的去囊衣带芯橘瓣分选

发布时间：2018-04-14 02:38

本文选题：图像处理 + 光谱　；参考：《农业工程学报》2017年09期

【摘要】：为实现去囊衣带芯橘瓣的机器视觉分选,给去囊衣带芯橘瓣的机器识别提供直接、准确的判别信号,采用荧光分光光度计对50颗去囊衣带芯橘瓣的橘芯和砂囊分别进行三维荧光光谱扫描,通过对橘芯、砂囊平均三维荧光光谱分析,确定了橘芯相对砂囊的特征荧光信号,据此对带芯橘瓣荧光图像识别的可行性进行了验证。检测发现,在370~390 nm紫外光激发下,橘芯和砂囊在440 nm处的荧光强度差异较大;橘芯与砂囊在370 nm紫外光激发下,440 nm处荧光强度分布的箱线图表明两者荧光强度分布存在明显差异,且以橘芯在440 nm处荧光强度的下四分位数(Q1)与砂囊在440 nm处荧光强度的上四分位数(Q3)的平均值作为分类标准,去囊衣带芯橘瓣检出准确率可达85%。对(370±2)nm激发下得到的单色(440±5)nm荧光图像进行二值化及形态学处理后,可在以砂囊为背景的橘瓣图形中显现橘芯亮斑。利用橘芯与砂囊的荧光特性差异进行机器视觉成像分析,可作为识别去囊衣带芯橘瓣的一种有效方法。
[Abstract]:In order to realize the machine vision sorting of the tangerine flap, it provides a direct and accurate discriminant signal for the machine recognition of the tangerine flap.A fluorescence spectrometer was used to scan 50 tangerine cores and sand sacs with a cored orange flap. The characteristic fluorescence signals of orange core relative to sand sac were determined by analyzing the average three dimensional fluorescence spectrum of orange core and sand sac.Based on this, the feasibility of fluorescence image recognition of orange flap with core is verified.It was found that the fluorescence intensity of orange core and sand sac was different at 440 nm under the excitation of 370 ~ 390nm ultraviolet light.The box diagram of fluorescence intensity distribution at 440 nm between orange core and sand sac under the excitation of 370 nm ultraviolet light shows that there is a significant difference in fluorescence intensity distribution between them.Using the average value of the lower quartile of fluorescence intensity at 440 nm and the upper quartile of fluorescence intensity of sand sac at 440 nm as the classification standard, the detection accuracy of the orange flap with the core of removing the capsule can reach 85%.After binarization and morphological processing of the monochromatic 5)nm fluorescence images excited by 370卤 2)nm, the orange core bright spots can be seen in the orange flaps with sand sac as the background.The difference of fluorescence characteristics between orange core and sand sac is analyzed by machine vision imaging, which can be used as an effective method to identify orange flaps with core.
【作者单位】：中国农业大学工学院;汇川盛业技术(北京)有限公司;
【基金】：国家重点研发计划(2016YFD0400305)
【分类号】：TP391.41;TS295.6

【相似文献】