用于手背静脉注射的图像自动识别和标注.pdfVIP

第37 卷 第9 期 红 外 技 术 Vol.37 No.9 2015 年9 月 Infrared Technology Sep. 2015 用于手背静脉注射的图像自动识别与标注 张 珣，郭永洪，李 刚，何金龙 （中国计量学院机电工程学院，浙江杭州 310018 ） 摘要：针对医疗领域的手背静脉识别研究甚少的问题，提出一种用于静脉注射的手背静脉自动识别的 方法。分析了静脉穿刺的特点以及适合注射的静脉条件，在近红外光下获得全局手背静脉图像的基础 上，首先对图像进行预处理获得完整的细化图像，再利用改进的区域生长法分别对每条静脉进行提取 和分析，最后用模板匹配法识别选择最适合穿刺的静脉部分。实验结果表明：该方法能够实现对适合 注射的静脉部分进行自动识别与标注。 关键词：静脉注射；特征提取；区域生长；模板匹配；静脉标注 中图分类号：TP391.41 文献标识码：A 文章编号：1001-8891(2015)09-0751-05 Image Automatic Recognition and Mark of Hand Vein Injection Parts ZHANG Xun ，GUO Yong-hong，LI Gang ，HE Jin-long (College of mechanical and electronic engineering, China Jiliang University, Hangzhou 310018, China) Abstract ：For the situation that there was little study on automatic recognition which was suitable for intravenous injection directly, we present a method that identifies the vein automatically. It analyzes the features of vein puncture and conditions suitable for intravenous injection. On the basis of getting global hand vein images under the near-infrared light, the improved region growing method is used to extract and analyze each of the vein. And template matching method is used to identify the most suitable part for vein puncturing. The test results show that the method can successfully identify and label the vein parts which are suitable for vein injection. Key words ：venous injection ，feature extraction，region growing ，template matching ，vein mark 高静脉穿刺的成功率，使一些病患家属经过简单静脉 0 引言