基于移动互联网和SVM技术的玉米病害识别技术研究

Research on Recognition of Maize Diseases Based on Mobile Internet and Support Vector Machine Technique

本文研究采用移动互联网技术中Web service的方式采集玉米病害的图片数据,使用移动终端作为病害信息采集的主要工具,在服务器端使用SVM(支持向量机)分类方法对病害加以识别,并将结果实时返回给移动终端。提高了农业生产者和技术人员的效率,同时证明SVM在分类识别时训练样本较少情况,仍然具备较好的泛化能力。实验中对SVM参数进行了调整,发现本文方法比较适合于玉米病害的识别。

This paper studied acquisition of image data of maize diseases using Web service method,which is a kind of mobile internet technique, using the mobile terminal as the main tool for disease information collection, identify maize disease by using SVM(support vector machine) classification method in server,then return the recognition results to the mobile terminal in real time. It improves the efficiency of agricultural producers and technical staff. At the same time, SVM classification still has good generalization ability with small training samples. In the experiment, SVM parameters were adjusted. Experimental results showed that this method is more suitable to the identification of maize diseases.