【摘要】 为了实现静脉滴注的自动测速,减轻医护人员的负担,本文将模板匹配算法运用到静脉注射滴速检测中.利用OpenCV对输液视频提取出每帧的图像序列,首先对视频图像利用混合高斯模型、积分投影算法、Hough变换进行液滴模板的提取,极大的避免了环境干扰对液滴模板的影响;然后利用模板匹配找出液滴帧并提取液滴帧号,最后根据帧号计算出输液速度.本文通过对50个静脉注射视频进行实验并与传统方法进行对比,结果表明该方法能够在各种复杂因素存在的情况下准确地对静脉注射滴速进行测量.更多还原

【Abstract】 In order to determine the speed of intravenous infusion automaticly and reduce the burden of medical staffs,this paper applies the template matching algorithm to the detection of infusion speed. The frame sequences are extracted from the infusion videos by using OpenCV,firstly,the template of video image is obtained by the mixed Gaussian model,the integral projection algorithm and the Hough transform can avoid the influence of environmental interference on the droplet template. Then,the droplet frame and its number are extracted by template matching,finally the infusion rate can be calculated based on the frame number. In this paper,50 intravenous video experiments were carried out. Being compared with the traditional method,the results showed that the modified method could effectively detect the drip rate in the presence of various complicated factors.更多还原