期刊文献+

轴承压印字符自动识别算法的研究 被引量:3

Study on Automatic Recognition Algorithm of Pressed Characters on Bearing
在线阅读 下载PDF
导出
摘要 为了方便轴承的生产和管理,针对轴承上字符的特点,提出了一种自动识别轴承压印字符的方法。首先借助环形光源,利用"反光差"产生并采集字符图像;然后利用随机Hough变换定位轴承圆心,并确定目标字符的位置;通过对单个字符进行仿射变换矫正字符区域;最后提取轮廓层次特征,并使用支持向量机识别。实验表明,该方法能够实现轴承字符的自动识别,并能满足实时性和识别率的要求。 In order to facilitate the production and management of bearing, for the characteristic of bearing characters, a method of automatic recognition of pressed characters on bearing is proposed. Character images are generated by the difference of reflectance with the help of the ring illumination. The bearing center is located using randomized Hough transform, and then characters is located. Each character region is redressed by affine transform. Contour features are extracted and characters are recognized by supporting vector machine. Experimental results show that the method presented in the paper can achieve automatic recognition of pressed characters on bearing and meet the real-time and recognition rate requirements.
出处 《电视技术》 北大核心 2012年第9期142-145,共4页 Video Engineering
基金 国家自然科学基金项目(60804013) "嵌入式机器视觉关键技术的研究与开发"产学研合作项目资助(011002)
关键词 字符识别 压印字符 随机HOUGH变换 仿射变换 character recognition pressed characters randomized Hough transform affine transform
  • 相关文献

参考文献8

二级参考文献36

共引文献52

同被引文献34

引证文献3

二级引证文献13

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部