CAJ | 학술논문

针对黄瓜常见叶部病斑图像的颜色特点，提出了将支持向量机（ Support Vector Machine ， SVM ）应用于黄瓜叶部病害识别中。首先，选择HSI 颜色系统作为图像特征提取的颜色空间，以减少光照强度对获取图像时的影响；然后，利用支持向量机进行叶部病害的识别。不同核函数的结果比较分析表明：径向基核函数对黄瓜叶部病害的识别率最高，最适于黄瓜霜霉病、角斑病和白粉病的分类识别；支持向量机识别方法在病害识别时训练样本少，具有很好的分类性能和泛化能力。
침대황과상견협부병반도상적안색특점，제출료장지지향량궤（ Support Vector Machine ， SVM ）응용우황과협부병해식별중。수선，선택HSI 안색계통작위도상특정제취적안색공간，이감소광조강도대획취도상시적영향；연후，이용지지향량궤진행협부병해적식별。불동핵함수적결과비교분석표명：경향기핵함수대황과협부병해적식별솔최고，최괄우황과상매병、각반병화백분병적분류식별；지지향량궤식별방법재병해식별시훈련양본소，구유흔호적분류성능화범화능력。
According to the color characteristics of cucumber common leaf disease image , it put Support Vector Machine ( SVM) forward for the recognition of cucumber leaf disease in this paper .First, it selected HSI color system as the color space for the image feature extraction in order to reduce the impact of light intensity for obtaining images ;Then , the pa-per used Support Vector Machine for the recognition of leaf disease .The experimental results show that: the results of comparative analysis of different kernel function demonstrated RBF kernel function get the highest recognition rate of cu -cumber leaf disease and is most suitable for the recognition of cucumber three kinds of disease .The classification method of SVM has good classification performance and generalization ability with a small sample of training in disease recogni -tion.