摘要
为快速识别芒果早期损伤,本文提出了基于高光谱图像特征和光谱特征的芒果损伤检测方法。图像特征检测方法总体检测准确率为98.22%;光谱特征检测方法总体检测准确率为98.67%。实验表明,利用高光谱图像特征和光谱特征可以对芒果早期损伤进行识别。
In order to quickly identify the early damage of mango, this paper proposes a mango damage detection method based on hyperspectral image features and spectral features. The overall detection accuracy of the image feature detection method was 98.22%;the overall detection accuracy of the spectral feature detection method was 98.67%. Experiments show that the early damage of mango can be identified by using hyperspectral image features and spectral features.
作者
李威
王晓辉
代作晓
LI Wei;WANG Xiaohui;DAI Zuoxiao(School of Automation Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Taicang Institute of Opto-Electronic Technology,Taicang 215400,China;Shanghai Institute of Technical Physics,Chinese Academy of Sciences,Shanghai 200083,China)
出处
《现代食品》
2022年第19期163-167,共5页
Modern Food
关键词
高光谱成像技术
芒果损伤
主成分分析
偏最小二乘法
hyperspectral imaging technology
mango damage
principal component analysis
partial least squares
