基于Kinect v2的呼吸运动监测方法研究

Research of respiratory motion monitoring using microsoft Kinect v2 sensor

为了提高基于Kinect相机的呼吸运动监测精度,提出了一种结合目标检测与目标跟踪算法的实时呼吸运动监测方法。该方法首先基于标记板形状特征利用霍夫圆检测算法进行目标检测,然后基于彩色图像结合Camshift目标跟踪算法实现单个或多个目标的跟踪,依据跟踪的结果获取目标位置的深度信息,并对获取的深度信息进行降噪处理。在此基础上,开发了一款基于Kinect相机的呼吸运动精准监测软件系统。实验结果表明,所提出的实时呼吸运动监测方法可有效提高呼吸运动监测的精度,与现有的基于深度图像获取呼吸运动数据方法相比,在体态偏移状态下数据采集准确度由47.9%提升至94.1%。同时,该方法具有良好的可视化效果。

In order to improve the accuracy of respiratory motion monitoring based on Kinect camera,a real-time respiratory motion monitoring method combining target detection and target tracking algorithm is proposed.Firstly,based on the shape features of the marker plate,the Hough algorithm for circle detection is used to detect the target,and then the color image combined with Camshift algorithm is used to track single or multiple targets.According to the tracking results,the depth information of the target position is obtained,and the obtained depth information is denoised.On this basis,a respiratory movement monitoring software system based on Kinect camera is developed.The experimental results show that the proposed real-time respiratory movement monitoring method can effectively improve the accuracy of respiratory movement monitoring.Compared with the existing methods based on depth image to obtain respiratory movement data,the accuracy of data acquisition is improved from 47.9%to 94.1%in the state of body posture deviation.At the same time,the method has good visualization effect.