基于距离场的非线性图像插值分割方法

Algorithm of Nonlinear Image Interpolation Segmentation Based on Distance Mask

对于医学序列切片,需要图像分割技术将图像中感兴趣区域ROI(region of interest)提取并进行三维重建.提出一种基于距离场的非线性图像插值分割方法,该方法克服了软组织因灰度值分布不均匀难分割的问题,可以快速提取ROI区域.实验表明用该方法进行医学序列切片的分割,可以得到良好的分割结果,并已成功地应用于医学数据的三维重建系统中.

In medical mierotomy, the ROl（region of interest） in image is needed to be picked up via image segmentation and to be 3D reconstructed. An algorithm of nonlinear image interpolation segmentation is thus presented on the based of distance mask to rise above effectively the difficulty of images of soft tissue because of the non-uniform distribution of gray level. The algorithm can quickly pick up ROI. The experiment showed that the algorithm can generate accurate result and it has successfully been applied to the 3D medical image visualization and reconstruction system.