黄桃碰伤和可溶性固形物高光谱成像无损检测

发布时间：2018-11-19 17:20

【摘要】：黄桃在线分级时,表面损伤和可溶性固形物同时在线检测。损伤和可溶性固形物是评价黄桃品质好坏的重要指标。采用高光谱成像技术,尝试对黄桃损伤和可溶性固形物进行同时检测。利用主成分分析法,首先对高光谱图像进行主成分分析得到最佳PC(principal component)图像,其次根据PC图像中各波长对其贡献率的大小确定最佳特征波长(550和720nm)并结合二值化,图像掩膜和阈值分割以及相关的图像处理技术对最佳光谱图像进行定性判别。其准确率最高达到94.6%,同时建立偏最小二乘定量回归模型对正常样品SSC(soluble solid content)含量进行预测,通过对模型的不断优化,实现了基于高光谱成像技术对黄桃碰伤和可溶性固形物同时检测。可溶性固形物分选准确率为79.2%。实验结果表明,利用高光谱成像技术可以实现对黄桃碰伤和可溶性固形物同时检测,该研究可以为实际在线分选提供理论依据和参考。
[Abstract]:The surface damage and soluble solids were detected at the same time when the yellow peach was classified on line. Damage and soluble solids are important indexes to evaluate the quality of yellow peach. Hyperspectral imaging was used to detect the damage and soluble solids of yellow peach simultaneously. Using principal component analysis (PCA), the best PC (principal component) image is obtained by principal component analysis (PCA), and the optimal characteristic wavelength (550 and 720nm) is determined according to the contribution rate of each wavelength in the PC image and combined with binarization. Image mask, threshold segmentation and related image processing techniques are used to qualitatively identify the best spectral images. At the same time, the partial least square quantitative regression model was established to predict the SSC (soluble solid content) content of normal samples. Based on hyperspectral imaging technology, the simultaneous detection of collision and soluble solids in yellow peach was realized. The resolution of soluble solids was 79.2%. The experimental results show that the hyperspectral imaging technique can be used to simultaneously detect the collision and soluble solids of yellow peach. This study can provide theoretical basis and reference for on-line sorting.
【作者单位】：华东交通大学机电与车辆工程学院光机电技术及应用研究所;
【基金】：国家“十二五”(863)计划项目(SS2012AA101306) 江西省优势科技创新团队建设计划项目(20153BCB24002) 南方山地果园智能化管理技术与装备协同创新中心(赣教高字[2014]60号) 江西省科技支撑计划(20121BBF60054)资助
【分类号】：O657.3;TS255.2

【相似文献】