基于支持向量机的葡萄病害图像识别方法

应用计算机图像处理技术和支持向量机识别方法研究了葡萄叶部病害的识别，以提高识别的准确性和效率。首先对采集到的葡萄病害彩色图像采用矢量中值滤波法去除噪声，然后采用统计模式识别方法和数学形态学对病叶图像进行了分割。最后提取了葡萄病叶彩色图像的纹理特征、病斑的形状特征和颜色特征，并用支持向量机的模式识别方法来识别葡萄病害。试验结果表明：支持向量机识别方法能获得比神经网络方法更好的识别性能；综合形状特征和纹理特征的支持向量机识别方法对葡萄病害的正确识别率优于只用形状特征或纹理特征的病种识别，综合颜色特征和纹理特征

Method for recognition of grape disease based on support vector machine

A new method for recognizing grape leaf disease by using computer image processing and Support Vector Machine(SVM) was studied to improve recognition accuracy and efficiency. At first, vector median filter was applied to remove noise of the acquired color images of grape leaf with disease. Then a method of statistic pattern recognition and mathematics morphology was introduced to segment images of grape leaf with disease. At last texture features, shape features and color features of color image of grape leaf with disease were extracted, and classification method of SVM for recognition of grape disease was used. Experimental results indicate that the classification performance of Support Vector Machine is better than that of neural networks. Recognition rate of grape disease based on SVM of shape and texture feature is better than that of only using the shape or texture feature, recognition rate of grape disease based on SVM of color and texture feature is higher than that of only using the color or texture feature.