高光谱成像技术结合集成学习的金钗石斛氮素检测

匡润^1,^2,^3,⁴, 龙腾^1,^3,⁴, 刘海林², 吴继辉^1,^3,⁴, 吕金胜^1,^3,⁴, 谢自然^1,^3,⁴, 刘文涛^1,^3,⁴, 兰玉彬^1,^3,⁴, 龙拥兵^1,^3,⁴, 王再花^2,^*, 赵静^1,^3,^4,^*

Hyperspectral Imaging Technology Combined With Ensemble Learning for Nitrogen Detection in Dendrobium nobile

KUANG Run^1,^2,^3,⁴, LONG Teng^1,^3,⁴, LIU Hai-lin², WU Ji-hui^1,^3,⁴, LÜ Jin-sheng^1,^3,⁴, XIE Zi-ran^1,^3,⁴, LIU Wen-tao^1,^3,⁴, LAN Yu-bin^1,^3,⁴, LONG Yong-bing^1,^3,⁴, WANG Zai-hua^2,^*, ZHAO Jing^1,^3,^4,^*

样品RGB图像和ROI区域的选择 (A): 样品RGB图像, (a): 正常; (b): 少数黄斑; (c): 黄斑数量较多; (d): 黄斑、褐斑; (e): 灰斑(B): ROI区域的选择过程, (a): 750 nm光谱图; (b): 背景反相; (c): 全局直方图均衡化; (d): 聚类分割; (e): 绿色主体和黄斑; (f): ROI区域