基于Kinect的ISS_CPD算法在混合现实医学中的应用

Application of Kinect-based ISS_CPD Algorithm in Mixed Reality Medicine

针对传统的图像匹配算法运用在混合现实医学中图像匹配速度慢、精准度低的问题,本文提出了一种基于ISS算法(内部形态描述子)和CPD算法(相干点漂移算法)的混合图像匹配方法,以实现三维图像与人体器官的快速匹配。首先,使用红外摄像头Kinect提取现实场景,生成点云图像,对点云图像进行去噪、平滑处理,优化图像质量;然后,使用ISS算法提取配准图像和参照图像的特征点,使用CPD算法进行快速配准,并使用点云三角化处理算法对配准图像进行三角化还原;最后,将算法进行封装,部署到混合现实眼镜Hololens中,从而实现混合现实医学图像实时配准。

In view of the slow speed and low accuracy of image matching in mixed reality medicine in traditional image matching algorithms,this paper proposes a hybrid image matching based on ISS algorithm(internal morphological descriptor)and CPD algorithm(coherent point shift algorithm)Method to achieve rapid matching of three-dimensional images with human organs.First,use the infrared camera Kinect to extract the real scene,generate a point cloud image,denoise and smooth the point cloud image,and optimize the image quality;then,use the ISS algorithm to extract the feature points of the registered image and the reference image,and use the CPD algorithm to perform Fast registration,and use the point cloud triangulation processing algorithm to triangulate the registered image;finally,the algorithm is packaged and deployed in the mixed reality glasses Hololens,so as to realize the real-time registration of mixed reality medical images.