光谱学与光谱分析, 2016, 36 (5): 1423, 网络出版: 2020-11-16  

基于高光谱图像多光谱参数的草莓成熟度识别

Identification of Strawberry Ripeness Based on Multispectral Indexes Extracted from Hyperspectral Images
作者单位
浙江大学生物系统工程与食品科学学院,浙江 杭州 310058
摘要
为了建立多光谱参数用于草莓成熟度的自动识别, 采用高光谱图像技术, 通过提取草莓样本ROI的平均光谱, 计算已有的八个成熟度参数Ind1, Ind2, Ind3, IAD, I1, I2, I3, I4的参数值, 并结合Fisher线性判别法判断八个参数对于三种成熟度(成熟、 接近成熟、 未成熟)草莓样本的分类识别效果, 发现基于I4参数的线性判别分析模型的识别效果最佳, 建模集和预测集识别准确率分别为90%和91.67%; 基于草莓样本的光谱特征, 提取与草莓成熟度相关的三个波长535, 675和980 nm, 并基于这三个波长和已有的参数形式, 构建了四个用于草莓成熟度检测的新参数: i1, i2, i3, i4, 通过Fisher线性判别法判断四个参数的分类识别效果, 发现基于参数i1, i2和i4的线性判别分析模型的识别效果均比参数I4好, 建模集和预测集识别准确率为95.83%, 95.83%, 95.83%和95%, 95%, 96.67%。 结果表明新建立的多光谱参数i1, i2和i4可以用于草莓成熟度的自动分类识别, 为草莓成熟度的在线检测提供了理论依据。
Abstract
In order to establish new multispectral indexes for automatic identification of strawberry ripeness, hyperspectral imaging technology was applied in this paper. Eight indexes: Ind1=R730+R640-2×R680, Ind2=R680/(R640+R730), Ind3=R675/R800, IAD=log10(R720/R670), I1=R650/R550, I2=R650/R450, I3= R650/(R450+R550), I4=2×R650-(R550+R450) were calculated by extracting average spectral of strawberry samples and their identification effects of strawberry samples in three ripening stages(mature, nearly mature and immature) were judged with Fisher linear discriminant(FLD). The result showed that the identification effects of linear discriminant analysis model based on index I4 was the best among 8 indexes and the identification accuracy of modeling and prediction set was 90% and 91. 67% respectively. Three wavelengths (535, 675, 980 nm) related to strawberry ripeness were extracted based on average spectral of strawberry samples and 4 new indexes were established based on these three wavelengths: i1=2×R675-(R980+R535), i2=R675/(R980+R535), i3=(R675-R535)/(R675+R535), i4=[R675-(R535+R980)]/[R675+(R535+R980)]. The identification effects was judged with FLD and the results showed that the effects of linear discriminant analysis models based on i1, i2, i4 were better than index I4 and the identification accuracy of modeling and prediction set was 95.83%, 95.83%, 95.83% and 95%, 95%, 96.67% respectively. In conclusion, new established indexes i1, i2, i4 could be used in automatic identification of strawberry ripeness.

蒋浩, 张初, 刘飞, 朱红艳, 何勇. 基于高光谱图像多光谱参数的草莓成熟度识别[J]. 光谱学与光谱分析, 2016, 36(5): 1423. Jiang Hao, Zhang Chu, Liu Fei, Zhu Hongyan, He Yong. Identification of Strawberry Ripeness Based on Multispectral Indexes Extracted from Hyperspectral Images[J]. Spectroscopy and Spectral Analysis, 2016, 36(5): 1423.

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!