基于改进的种子生长算法辅助静脉穿刺的研究

Research on vein puncture assisted by improved seed growth algorithm

针对目前临床手臂静脉辅助穿刺研究较少的情况,提出一种改进种子算法提高识别效率。首先通过红外摄像头获取静脉图像进行预处理操作,提取特征并二值化,然后进行细化操作并获得所有分支点和端点,在此基础上利用改进的种子生长法对细化图像标号和剪裁,最后通过斜率等条件建立标准模板进行模板匹配,从而提取出最佳穿刺部位。实验结果显示该方法能够对适合穿刺的静脉部位实现较高识别率。

Because of less research on the arm vein assisted puncture in the current medical field,to propose an improved seed algorithm to improve the recognition efficiency.Firstly,the infrared camera acquired the vein image for preprocessing,extracted features and binarizes,then refined the operation and obtained all the branch points and endpoints.On this basis,the improved seed growing method was used to refine the image marking and clipping.Finally,a standard template was established by slope and other conditions for template matching to extract the optimal puncture site.The experimental results show that the method can achieve high recognition for the venous site suitable for puncture.