摘要
血管壁超声图像边缘检测是IVUS图像处理中必不可少的步骤和重要的环节。传统的IVUS图像边缘检测方法是直接采用Snake动态轮廓模型、GVF Snake动态轮廓模型对IVUS图像进行处理,这些方法不易得到确切的血管壁边缘边缘轮廓。本文采用先对IVUS原图像进行预处理的方法。在预处理中,提出了采用不同高斯滤波方差下的Canny算子边缘检测和Hough变换圆检测方法。之后再采取GVF变换、图像除噪与图像迭加等处理方法。实验结果验证了该IVUS图像处理方法的有效性。
The detection of the media-adventitial borders is a key step in intravascular ultrasound （IVUS） processing. The traditional method for IVUS is based on the Snake model or GVF Snake model. Actually, it is not easy to get the precise media-adventitial border of the IVUS image by only using the one of the above method. In this paper,at first, the IVUS original image is preprocessed with the Canny edge detection of the Gaussian filters in different variances and Circle Hough Transform. Then the preprocessed image is transformed by GVF. Finally, the noise dots on the GVF transfor ming image are removed and the desirable media-adventitial borders of IVUS image are obtained, The experimental results confirm the validity of the new method.
出处
《国外电子测量技术》
2008年第9期14-16,20,共4页
Foreign Electronic Measurement Technology
