光谱学与光谱分析
光譜學與光譜分析
광보학여광보분석
SPECTROSCOPY AND SPECTRAL ANALYSIS
2014年
5期
1362-1366
,共5页
程术希%谢传奇%王巧男%何勇%邵咏妮
程術希%謝傳奇%王巧男%何勇%邵詠妮
정술희%사전기%왕교남%하용%소영니
高光谱成像%特征波长%线性判别分析%最小二乘-支持向量机%番茄%早疫病
高光譜成像%特徵波長%線性判彆分析%最小二乘-支持嚮量機%番茄%早疫病
고광보성상%특정파장%선성판별분석%최소이승-지지향량궤%번가%조역병
Hyperspectral imaging%Effective wavelengths (EW )%Linear discriminant analysis (LDA )%Least square-support vector machines (LS-SVM )%Tomato%Early blight
提出了基于连续投影算法（successive projections algorithm ，SPA）、载荷系数法（x-loading weights ， x-LW）和格拉姆-施密特正交（gram-schmidt orthogonalization ，GSO）提取特征波长的高光谱成像技术检测番茄叶片早疫病的方法。首先获取380～1023 nm波段范围内70个健康和70个染病番茄叶片的高光谱图像信息，然后提取健康和染病叶片感兴趣区域（region of interest ，ROI）的光谱反射率值，建立番茄叶片早疫病的最小二乘-支持向量机（least squares-support vector machine ，LS-SVM ）鉴别模型，建模集和预测集的鉴别率都是100％。再通过SPA 、x-LW和GSO提取特征波长（effective wavelengths ，EW），并建立EW-LS-SVM和特征波长-线性判别分析（ew-linear discriminant analysis ，EW-LDA ）鉴别模型。结果显示，每个模型的鉴别效果都很好，EW-LS-SVM模型中预测集的鉴别率都达到了100％，EW-LDA模型中预测集的鉴别率分别是100％，100％和97.83％。基于 SPA ， x-LW 和 GSO 所建模型的输入变量分别是4个（492，550，633和680nm），3个（631，719和747 nm）和2个（533和657 nm），较少的特征波长便于实时检测仪器的开发。结果表明，高光谱成像技术检测番茄叶片早疫病是可行的，SPA ，x-LW和GSO都是非常有效的特征波长提取方法。
提齣瞭基于連續投影算法（successive projections algorithm ，SPA）、載荷繫數法（x-loading weights ， x-LW）和格拉姆-施密特正交（gram-schmidt orthogonalization ，GSO）提取特徵波長的高光譜成像技術檢測番茄葉片早疫病的方法。首先穫取380～1023 nm波段範圍內70箇健康和70箇染病番茄葉片的高光譜圖像信息，然後提取健康和染病葉片感興趣區域（region of interest ，ROI）的光譜反射率值，建立番茄葉片早疫病的最小二乘-支持嚮量機（least squares-support vector machine ，LS-SVM ）鑒彆模型，建模集和預測集的鑒彆率都是100％。再通過SPA 、x-LW和GSO提取特徵波長（effective wavelengths ，EW），併建立EW-LS-SVM和特徵波長-線性判彆分析（ew-linear discriminant analysis ，EW-LDA ）鑒彆模型。結果顯示，每箇模型的鑒彆效果都很好，EW-LS-SVM模型中預測集的鑒彆率都達到瞭100％，EW-LDA模型中預測集的鑒彆率分彆是100％，100％和97.83％。基于 SPA ， x-LW 和 GSO 所建模型的輸入變量分彆是4箇（492，550，633和680nm），3箇（631，719和747 nm）和2箇（533和657 nm），較少的特徵波長便于實時檢測儀器的開髮。結果錶明，高光譜成像技術檢測番茄葉片早疫病是可行的，SPA ，x-LW和GSO都是非常有效的特徵波長提取方法。
제출료기우련속투영산법（successive projections algorithm ，SPA）、재하계수법（x-loading weights ， x-LW）화격랍모-시밀특정교（gram-schmidt orthogonalization ，GSO）제취특정파장적고광보성상기술검측번가협편조역병적방법。수선획취380～1023 nm파단범위내70개건강화70개염병번가협편적고광보도상신식，연후제취건강화염병협편감흥취구역（region of interest ，ROI）적광보반사솔치，건립번가협편조역병적최소이승-지지향량궤（least squares-support vector machine ，LS-SVM ）감별모형，건모집화예측집적감별솔도시100％。재통과SPA 、x-LW화GSO제취특정파장（effective wavelengths ，EW），병건립EW-LS-SVM화특정파장-선성판별분석（ew-linear discriminant analysis ，EW-LDA ）감별모형。결과현시，매개모형적감별효과도흔호，EW-LS-SVM모형중예측집적감별솔도체도료100％，EW-LDA모형중예측집적감별솔분별시100％，100％화97.83％。기우 SPA ， x-LW 화 GSO 소건모형적수입변량분별시4개（492，550，633화680nm），3개（631，719화747 nm）화2개（533화657 nm），교소적특정파장편우실시검측의기적개발。결과표명，고광보성상기술검측번가협편조역병시가행적，SPA ，x-LW화GSO도시비상유효적특정파장제취방법。
Identification of early blight on tomato leaves by using hyperspectral imaging technique based on different effective wavelengths selection methods (successive projections algorithm ,SPA ;x-loading weights ,x-LW ;gram-schmidt orthogonaliza-tion ,GSO) was studied in the present paper .Hyperspectral images of seventy healthy and seventy infected tomato leaves were obtained by hyperspectral imaging system across the wavelength range of 380~1023 nm .Reflectance of all pixels in region of in-terest (ROI) was extracted by ENVI 4 .7 software .Least squares-support vector machine (LS-SVM ) model was established based on the full spectral wavelengths .It obtained an excellent result with the highest identification accuracy (100% ) in both calibration and prediction sets .Then ,EW-LS-SVM and EW-LDA models were established based on the selected wavelengths suggested by SPA ,x-LW and GSO ,respectively .The results showed that all of the EW-LS-SVM and EW-LDA models per-formed well with the identification accuracy of 100% in EW-LS-SVM model and 100% ,100% and 97.83% in EW-LDA model , respectively .Moreover ,the number of input wavelengths of SPA-LS-SVM , x-LW-LS-SVM and GSO-LS-SVM models were four (492 ,550 ,633 and 680 nm) ,three (631 ,719 and 747 nm) and two (533 and 657 nm) ,respectively .Fewer input variables were beneficial for the development of identification instrument .It demonstrated that it is feasible to identify early blight on to-mato leaves by using hyperspectral imaging ,and SPA ,x-LW and GSO were effective wavelengths selection methods .
