发光学报, 2019, 40 (6): 808, 网络出版: 2019-09-03   

基于近红外的柚子品种判别和糖度检测通用模型

A General Model for Judging and Brix Detection of Grapefruit Variety Based on Near Infrared
作者单位
华东交通大学 机电与车辆工程学院, 江西 南昌 330013
摘要
针对厚皮果透光性差、不同柚子品种糖度在线检测要单独建模等问题, 本文以柚子为研究对象, 采集有效光谱325条, 对比分析不同柚子品种在710 nm和800 nm附近的两个吸收峰光谱响应特性。550~970 nm全波段范围内的光谱采用SPA无信息消除后建立柚子偏最小二乘判别模型误判率为1.25%; 偏最小二乘法在550~970 nm全波段范围和去差异化后750~930 nm波段范围的预测相关系数分别为0.58和0.86, 预测均方根误差RMSEP分别为0.84和0.55。实验结果表明, 连续投影法结合偏最小二乘判别模型可以实现不同柚子品种的定性判别, 变异系数法对光谱去差异化后结合最小二乘模型对厚皮果柚子可溶性固形物的定量检测效果最佳, 该研究为不同品种的厚皮果在线分选技术提供了参考和理论依据。
Abstract
In view of the poor light transmission of thick skin fruit and the independent modeling of the sugar content of different grapefruit varieties, the grapefruit was the research object, and 325 effective spectra were collected to compare and analyze the spectral responses of two absorption peaks of different grapefruit varieties near 710 nm and 800 nm. In the 550-970 nm full-band spectrum, the false positive rate of the grapefruit partial least squares discriminant model is 1.25% after SPA no information elimination; the partial least squares method is in the 550-970 nm full-band range and the de-differential 750-930 nm band. The prediction correlation coefficients of the range are 0.58 and 0.86, respectively, and the predicted root mean square error RMSEP is 0.84 and 0.55, respectively. Experiments show that the continuous projection method combined with the partial least squares discriminant model can realize the qualitative discrimination of different grapefruit varieties. The coefficient of variation method is the best to quantify the soluble solids of the thick fruit grapefruit with the least squares model. This study provides a reference and theoretical basis for the online sorting technology of different varieties of thick fruit.

李雄, 刘燕德, 欧阳爱国, 孙旭东, 胡军, 姜小刚, 欧阳玉平. 基于近红外的柚子品种判别和糖度检测通用模型[J]. 发光学报, 2019, 40(6): 808. LI Xiong, LIU Yan-de, OUYANG Ai-guo, SUN Xu-dong, HU Jun, JIANG Xiao-gang, OUYANG Yu-ping. A General Model for Judging and Brix Detection of Grapefruit Variety Based on Near Infrared[J]. Chinese Journal of Luminescence, 2019, 40(6): 808.

本文已被 1 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!