多光谱成像技术诊断植物病虫害的人工神经网络模型.PDFVIP

34 5 V ol. 34 No. 5 2 0 0 8 9 OP T ICA L T E CHN IQ U E Sep . 200 8 : 100215 82 ( 200 8) 多光谱成像技术诊断植物病虫 害的人工神经网络模型 1 2 1 2 1 3 4 冯洁 廖宁放 赵波 罗永道 李宝聚 戴志福 ( 1. 1000 81; 2 . 650092) ( 3. 10008 1; 4. 661100) : 14 BP 14 850nm : ; BP ; : O432. 4 : A Multispectral imaging techniques diagnosing plant diseases and insect pests using artificial neural networks 1, 2 1 2 1 1 3 4 FENG Jie , LIAO Ningfang , ZHAO Bo , LIANG Minyong , LUO Yongdao , LI Baoj u , DAI Zhifu ( 1. N at ional L aboratory of Colour Science and Eng ineering Beijing I nstitute of T echnolog y Beij ing 1000 81 China) ( 2. Yunnan N orm al U niv ersity Kunm ing 650092 China) ( 3 . T he Chinese Academy of Agr icultural Sciences Beij ing 1000 81 China) ( 4. Honghe U niversit y Y unnan 661100 China) Abstract: For a reliable diagnosis of plant diseases and in sect pest s art if icial neural netw or ks and mutisp ectral im aging technique are proposed to diagnosise t hree cucumbers diseases. In the ex perim ent the cucumbers multisp ectral im ag es of 14 vis ible lights ch annels near infrared ch annel and panchrom atic ch annel are captured using narro wband mult ispectral im aging sys tem and classified by the BP neural netw or k. T he coefficients o f output and prediction are obtained . T he result show s that the m