MRI中的肿瘤边缘的自动检测

MRI用于临床诊治时，肿瘤大小的自动检测一直是尚未圆满解决的问题，其中的关键技术是肿瘤边缘的自动检测。边缘检测是计算机初级视觉中的一个难题，一直受到图像处理研究人员的关注。虽然边缘检测的算子很多，但都是在噪声滤除与边缘检出之间作权衡取舍，本研究中提出了一种结合形态滤波器和广义模糊算子自动检测MRI中的肿瘤边缘的新方法，它能在保留边缘细节的同时最大程度地抑制噪声，实验结果表明，在噪声抑制和边缘点检出的折衷中，该方法具有比传统边缘缘检测算子更好的表现。

Automatic Edge_Detection of Tumor in MRI

Measuring the tumors in MRI automatically has always been a problem and no satisfactory solution has yet been offered. The key technology is the automatic detection of the tumor edges. Edge_detection is also a big problem in computational low_level vision and has always attracted the attention of image_processing researchers. There are many edge_detectors, but they are limited in making a balance between noise_suppressing and edge_detection. This paper presents a new method, which detects the edges of tumor in MRI using morphological filter and generalized fuzzy operator. It can preserve edge details while suppressing the noises. Experimental result shows that the new method does better than traditional methods.