光谱学与光谱分析
光譜學與光譜分析
광보학여광보분석
SPECTROSCOPY AND SPECTRAL ANALYSIS
2014年
8期
2225-2228
,共4页
宋革联%余俊霖%刘飞%何勇%陈丹%莫旺成
宋革聯%餘俊霖%劉飛%何勇%陳丹%莫旺成
송혁련%여준림%류비%하용%진단%막왕성
虫害%生命状态%高光谱成像%快速检测
蟲害%生命狀態%高光譜成像%快速檢測
충해%생명상태%고광보성상%쾌속검측
Insect pest%Life state%Hyperspectral imaging%Fast detection
采用近红外高光谱成像技术对菜青虫的存活与死亡状态进行了研究,通过提取菜青虫不同状态的光谱信息,建立判别分析模型。以不同预处理方法对所提取的951.5~1649.2 nm光谱进行预处理,并建立偏最小二乘判别分析(partial least square-discriminant analysis,PLS-DA)模型对菜青虫的生死状态进行判别分析,判别正确率接近或达到100%。用移动平均(moving average,MA)5点平滑光谱分别采用连续投影算法(successive projections algorithm,SPA)以及加权回归系数(weighted regression coefficient,Bw )分别选取了17和20个特征波长进行生与死状态的判别。基于特征波长建立了 PLS-DA,K 最邻近节点算法(K-nea-rest neighbor,KNN),BP神经网络(back propagation neural network,BPNN)以及支持向量机(support vec-tor machine,SVM)模型,判别正确率接近100%。结果表明采用近红外高光谱成像技术对菜青虫生命状态的研究是可行的,为作物虫害的快速诊断提供了新方法。
採用近紅外高光譜成像技術對菜青蟲的存活與死亡狀態進行瞭研究,通過提取菜青蟲不同狀態的光譜信息,建立判彆分析模型。以不同預處理方法對所提取的951.5~1649.2 nm光譜進行預處理,併建立偏最小二乘判彆分析(partial least square-discriminant analysis,PLS-DA)模型對菜青蟲的生死狀態進行判彆分析,判彆正確率接近或達到100%。用移動平均(moving average,MA)5點平滑光譜分彆採用連續投影算法(successive projections algorithm,SPA)以及加權迴歸繫數(weighted regression coefficient,Bw )分彆選取瞭17和20箇特徵波長進行生與死狀態的判彆。基于特徵波長建立瞭 PLS-DA,K 最鄰近節點算法(K-nea-rest neighbor,KNN),BP神經網絡(back propagation neural network,BPNN)以及支持嚮量機(support vec-tor machine,SVM)模型,判彆正確率接近100%。結果錶明採用近紅外高光譜成像技術對菜青蟲生命狀態的研究是可行的,為作物蟲害的快速診斷提供瞭新方法。
채용근홍외고광보성상기술대채청충적존활여사망상태진행료연구,통과제취채청충불동상태적광보신식,건립판별분석모형。이불동예처리방법대소제취적951.5~1649.2 nm광보진행예처리,병건립편최소이승판별분석(partial least square-discriminant analysis,PLS-DA)모형대채청충적생사상태진행판별분석,판별정학솔접근혹체도100%。용이동평균(moving average,MA)5점평활광보분별채용련속투영산법(successive projections algorithm,SPA)이급가권회귀계수(weighted regression coefficient,Bw )분별선취료17화20개특정파장진행생여사상태적판별。기우특정파장건립료 PLS-DA,K 최린근절점산법(K-nea-rest neighbor,KNN),BP신경망락(back propagation neural network,BPNN)이급지지향량궤(support vec-tor machine,SVM)모형,판별정학솔접근100%。결과표명채용근홍외고광보성상기술대채청충생명상태적연구시가행적,위작물충해적쾌속진단제공료신방법。
Near-infrared hypserspectral imaging technology was applied for the discrimination of a variety of life states,the judg-ment of being alive or death.Discrimination models were built based on spectral data of pieris rapaes acquired during different life states.The wavelengths from 951. 5 to 1 649. 2 nm were used for analysis after the removal of spectral region with obvious noi-ses at the beginning and the end.And the spectra data of 951. 5~1 649. 2 nm were preprocessed by different pretreatment meth-ods.To discriminate the state of being alive or death of pieris rapaes,discrimination models were built based on the spectral data processed by different pretreatment methods.Results showed that the discriminant accuracy can approach or attain 100%.Thus the method was proved to be useful for the discrimination of the state of being alive or death of pieris rapaes.After the spectral data were preprocessed by moving average (MA)algorithm,17 characteristic wavelengths were extracted based on weighted re-gression coefficient (Bw)and 20 were extracted based on successive projections algorithm (SPA)to identify the state of being a-live or death of pieris rapaes.Four classification methods based on characteristic wavelengths,including partial least squares-dis-criminant analysis (PLS-DA),K-nearest neighbor algorithm (KNN),back propagation neural network (BPNN)and support vector machine (SVM)were used to build discriminant models for identifying the state of being alive or death of pieris rapaes. The discriminant accuracy all can approach or attain 100%.