支持向量机在斑潜蝇虫害叶片光谱分析中的应用  

Classification Visible Spectra of Leafminer-infected Leaves by Support Vector Machine

作　　者：吴达科[1] 马承伟[2] 杜尚丰[3] 

机构地区：[1]西南大学工程技术学院 [2]中国农业大学农业部设施农业生物环境工程重点开放实验室 [3]中国农业大学信息与电气工程学院

出　　处：《农业机械学报》2007年第10期87-90,共4页Transactions of the Chinese Society for Agricultural Machinery

基　　金：国家自然科学基金资助项目(项目编号:60374030);北京市教育委员会共建项目建设计划资助(项目编号:XK100190550);西南大学博士启动基金资助(项目编号:SWUB2006015)

摘　　要：测量了斑潜蝇虫害叶片的可见光反射光谱。运用图像处理技术计算出虫害叶片的破损率;采用光谱分析方法选出对叶片破损程度敏感的波长点。运用支持向量机(SVM)和多光谱法,建立基于敏感波长的受害叶片反射光谱分类模型并进行分类试验。试验结果表明,多光谱法分类精度为90%,SVM为93.8%(多项式核函数)和96.9%(RBF核函数),高于多光谱法。Leafminer is a vegetable insect pest. The visible spectral reflectance of the leafminer （Liriomyza sativae Blanchard） infected plant leaves was measured. Damaged degrees （DD） of leaves were worked out with the image processing technology, and the sensitive wavelengths related to them were also selected via spectroscopy analysis. Using the support vector machine （SVM） and the multi-spectral method, the spectral classifying experiments were done and the classifying models were set up to recognize the leaf spectra with different DD. The results that used the SVM and the multi-spectral methods were contrasted. The results showed that the classifying precisions of the SVM are 93.8% （using the polynomial-based kernel function） and 96.9%（using the RBF kernel function）, which excelled to the multi-spectral method （whose classifying precision is 90%）.

关 键 词：叶片 可见反射光谱 图像处理 美洲斑潜蝇 支持向量机 

分 类 号：S431.9[农业科学—农业昆虫与害虫防治]