摘要
Mandelbrot 的分形理论认为,自然界中的大量图像都可以用分形布朗运动模型来描述,而分形维数可以表征图像自相似性和灰度表面粗糙度。充分地利用图像的分形维数的特性和小波变换的边缘保持好、适合人眼的特点来增强图像,辅助医学临床诊断。通过对X 光胸片图像的增强实验表明:该方法不仅能增强图像中较不敏感的细节内容,还能增强视觉上很难感受到的差异结构。
According to mandelbrot's theory, the nature images can be regarded as the result of FBM (fractional brownian motion) and fractal dimension can describe excellently the character of image self-similitude and coarseness of its gray surface. A method of image enchantment is proposed based on fractal & wavelet theory which can be used to assist clinic medical diagnosis. Experiments on x-ray sternum images show that the method can strengthen the non- sensitive parts in image and differential structures which is hard to distin- guish for human vision.
出处
《计算机工程与设计》
CSCD
北大核心
2005年第3期807-809,共3页
Computer Engineering and Design
