摘要
为了提高仓库管理系统的性能,将支持向量机用于产品编号的模式识别。采用基于投影法的图像处理算法提取编号数字;对倾斜的数字进行矫正,并对提取的数字进行归一化;构造支持向量机分类器对归一化的数字进行识别;通过对一组数字样本的测试,分析了支持向量机参数与分类器的识别率的关系。测试结果表明,支持向量机分类器可以在小样本的情况下获得较高的识别率。
In order to improve the performance of storage management systems, support vector machine （SVM） was used in the pattern recogni- tion of products serial numbers. Digits of products serial numbers were extracted through a series of image process algorithms based on projection; the skew digits were rectified, and then all extracted digits were normalized;an SVM classifier was crea- ted to recognize the normalized digits; through a test of a group of samples, the correlation between the SVM parameters and the recognition ratio of the classifier was analyzed. The results of the test showed that the SVM classifier could obtain high recognition ratio in the case of a small sample.
出处
《机械与电子》
2012年第2期63-66,共4页
Machinery & Electronics
关键词
支持向量机(SVM)
数字提取
数字识别
识别率
support vector machine （SVM）
digit extraction
digit recognition
recognition ratio
