针对医学图像组织间不明显现象,提出了一种基于模糊规则和小波变换的医学图像锐化增强算法(MFRWTE)。为了避免过增强现象和放大噪声,对不同尺度的小波系数进行锐化增强时,首先计算该尺度低频系数中心像素与其邻域像素的相容性,利用模糊规则将像素分为低细节,中细节和高细节三类,然后利用自适应算法计算非线性细节增益系数。最后通过把增益系数与细节小波系数相乘,小波重建后得到增强图像。实验结果表明,提出的算法对图像细节进行增强的同时能够有效地抑制噪声。用户也可以根据图像的特征,方便的通过调节中细节区域增强因子或小波分解层数获得满意的增强效果。
In order to cope with the blurred phenomena between tissues in medical images, a medical image sharpening algorithm based on fuzzy rules and wavelet transform was proposed. When detail coefficients of different wavelet layers were enhanced, each pixel was fuzzily grouped into one of three regions: low detail region, middle detail region and high detail region. Then enhancement factor was evaluated according to the local feature. At last, the enhanced image was obtained using inverse wavelet transforms. Experimental results showed that the proposed algorithm could enhance image details while restraining noises. Meanwhile, the users ean obtain satisfactory enhancement result conveniently by adjusting the enhancement factor of middle detail region or wavelet decomposition levels according to the characters of the image.