基于像素灰阶熵的自适应增强算法在乳腺CR图像中的应用

The adaptive enhancement means based on gray entropy applied in mammary gland CR image

乳腺完全是由密度接近软组织构成,因此乳腺CR(Computed Radiography)放射图像对比度小,轻微的差别变化都可能是肿瘤的表现,要对其进行增强处理方能满足医生临床诊断的需要。而目前通用的乳腺CR图像增强算法对比度和噪声增强过度,丢失细节,为此提出了采用一种基于像素灰阶熵的自适应边缘增强算法对乳腺CR图像进行增强。该算法使用模糊影像增加对选择空间频率的响应,以增强乳腺CR图像结构边缘和细节;算法能根据乳腺CR图像灰度特性即像素灰阶熵来自适应的调节增强程度的加权因数K。实验证明,该算法处理后的乳腺CR图像细节丰富,信噪比高,增强后的图像具有良好的视觉效果,该算法是一种有效的适合乳腺CR医学图像的边缘细节增强算法。

Mammary gland is composed of similar density soft tissue.Mammary gland CR image is bad contrast.Small difference vary is the possibility for checking tumour,so it is necessary to enhance mammary gland CR image to improve its visual quality in order to satisfy doctor diagnosis.However presently common enhancement algorithms,the contrast and noise is enhanced overly and image details are lost.Aiming at the defects,mammary gland CR medicine image adaptive enhancement arithmetic is put forward based on image gray entropy.The arithmetic adapts the unsharp masking to enhance select frequency response in order to enhance mammary gland CR image edge details.It can adjust factor K based on image gray characteristic.Experiment results demonstrate that mammary gland CR image enhanced by the algorithm has abundant detail and high signal to noise ratio.CR image enhanced has good visual effect.The method is fit for edge detail enhancement of mammary gland CR medicine image.