Research On ROI Segmentation And 3D Reconstruction Algorithm

In this paper, visualization of medical image sequences of the two basic issues have been conducted in-depth study, namely, medical image sequence region of interest (Region of Interest, ROI) segmentation and visualization of three-dimensional (3D Visualization) Theory and Methods. This article firstly studies a variety of ROI segmentation method of image sequences, proposed by combining pulse coupled neural network model and active contour model to achieve the method of the automatic segmentation of medical image sequences, and then base on the results of segmentation using marching cubes and contour tiling surface rendering three-dimensional reconstruction method to construct the interested object, output the three-dimensional image, But the surface of three-dimensional image is not smooth, the shadow of a series of cascades make the surface rough, finally using based contour surface of the vector field smoothing method to smooth the surface of three-dimensional images, making the final display-result is more close to the situation in the real.