CT 图象三维重建过程中一种改进的边缘提取法

Improved Edge Detection Method for CT Image 3D Reconstruction

采用将最佳阈值搜寻范围限定在一定的灰度阈值范围内来保证类间方差为单峰，再利用类间方差测度准则，求取图象的灰度阈值、梯度阈值，利用这两个阈值将ＣＴ图象中各器官的边缘提取出来，作为三维重建的轮廓输入．结果表明，这种方法有较好的实时性和适应性．

For Reddi method does well when the between class variance presents a single peak,the seeking range of optimum threshold is confined to make the between class variance be single peak.Then,the image gray value threshold is acquired by use of the between class variance measure and the image gradient threshold.Through those two thresholds,the edges of every organ in the CT image are detected to be used as the outline input for 3d reconstruction.The result compared with the other traditional methods is given.The experiments show that this method has good reality and feasibility.