医学图象的边界处理

医学图象质量的好坏直接影响医务人员对疾病的诊断与治疗。由于有时候医学图象的边界不清晰，甚至很模糊，人眼难以判断，因此需要一种特殊的算法来实现边界点的检测，边界点检测的方法很多，作者讨论用微分法借助于计算机来处理医学图象的边界问题。利用微分运算获得的梯度算子。根据区域边界处灰度呈现不连续的特点，通过对图象函数ｆ（ｘ，ｒ）（灰度特性函数）进行必要的运算，来达到检测边界元素的目的，进而实现边界点的连接，

The Boundary Processing of Medical Image

A good image is important for diagnosis and treatment of disease Because medical image borders are often not clearly defined,and sometimes difficult for human eyes to detect,so special algorithm is needed to obtain accurate boundary spots. There are many methods for boundary detection . Here we discuss how to deal with the boundary problem of medical images by differential calculus with the computer.With the gradient operator produced by differential calculus. and by the nature of noncontinuity of boundary gray levels,we can detect the boundary pixels by special calculus, and concatenate them one by one , then a closed contour curve can be formed. This com- puter program has self-adapting and self- learninge performance. lt is easy to usu.