摘要
根据日光温室黄瓜病害的彩色纹理图像的特点,将支持向量机和色度矩方法用于识别黄瓜病害。在进行分类时,首先以色度矩作为特征向量,然后将支持向量机分类方法应用于黄瓜病害的识别。通过黄瓜病害纹理图像识别实验分析表明,以色度矩作为病害彩色纹理图像的特征向量简便、快捷、分类效果好;支持向量机分类方法在黄瓜病害训练样本较少时也具有良好的分类能力和泛化能力,非常适合于黄瓜病害的分类问题。不同分类核函数的相互比较分析表明,线性核函数最适合黄瓜病害识别。
According to the features of color texture image of cucumber disease in sunlight greenhouse, recognition of cucumber disease using Support Vector Machine（SVM） and chromaticity moments is introduced. At first, extracting features of chromaticity moments is done, then classification method of SVM for recognition of cucumber disease is discussed. Experimentation with cucumber disease is conducted and the results prove that chromaticity moments are simple, efficient, and effective for recognition of cucumber disease image. SVM method has excellent classification and generalization ability in solving learning problem with small training set of sample, and is fit for classification of cucumber disease. The comparison of different kernel functions for SVM shows that liner kernel function is most suitable for recognition of cucumber disease.
出处
《农机化研究》
北大核心
2006年第2期151-153,160,共4页
Journal of Agricultural Mechanization Research
关键词
农业基础科学
黄瓜病害
理论研究
图像处理
日光温室
支持向量机
色度矩
agricultural basic science
cucumber disease
theoretical research
image processing
sunlight greenhouse
support vector machine
chromaticity moments
