【摘要】 为了解决人工鉴别水稻指纹图谱效率低下的问题,提出一种计算机辅助鉴别方法。提取用SSR分子标记技术构建的水稻DNA图谱条带轮廓,计算各条带区域的质心,将条带质心转换形成特征向量,并利用支持向量机构造出品种分类器,从而实现对未知样品的自动分类。该方法利用支持向量机较强的泛化能力,解决特征提取后不同品种样本之间细微差别所带来的识别困难。实验结果表明,该方法识别率总体上比欧几里得距离分类要高,可以辅助提高水稻品种鉴别效率。更多还原

【Abstract】 To solve the problem of low efficiency of artificial identification to paddy’s seeds,a method of computer-aided identification is proposed.The procedure of the feature vector by extracting each region outline from the fingerprint of a paddy’seed with SSR molecular marker technology is provided.An algorithm of classifier based on SVM is constructed.The experimental results show this method is feasible,which is compared to the method based on Euclidean distance,indicated that this algorithm is better,and can help to improve the efficiency of variety identification on paddy’s seeds.更多还原