基于MATLAB的乳腺钼靶图像直方图增强方法的比较

Comparative Study of Approaches of Histogram Enhancement to Mammography Image Based on MATLAB

目的比较不同直方图增强方法对改善乳腺钼靶图像质量的效果。方法基于MATLAB编程,分别采用直方图均衡化(HE)、直方图规定化(HS)和对比度受限自适应直方图均衡化(CLAHE)对乳腺钼靶图像进行增强处理;利用峰值信噪比(PSNR)客观评价图像的噪声水平。结果HE对图像对比度的增强效果一般,图像细节反而有所下降;HS可选择匹配直方图函数的类型,从而选择性增强所需灰度范围,但图像噪声水平在三种方法中最高;CLAHE能很好地增强图像各部分的对比度,且图像噪声水平最低。结论应用直方图增强方法处理乳腺钼靶图像,在图像噪声水平和灰度增强上,CLAHE明显优于HE和HS。

Object To compare the effects on improving the quality of mammography image among different histogram enhancement algorithms. Methods： Three processing algorithms, histogram equalization （HE）, histogram specification（HS） and contrast limited adaptive histogram equalization（CLAHE）, were applied to enhance a Inammography image based on programming software MATLAB, then evaluated the image and evaluated the level of noise objectively by means of the PSNR（peak signal to noise ratio）. Results： HE can not enhance the contrast of whole image obviously, at the same time, it can decline the details of whole image; HS can choose the type of histogram function to match, so it can enhance the contrast of whole image selectively, but there is the highest level of noise among these three algorithms; CLAHE can enhance the contrast of whole image obviously, and the level of noise is the lowest. Conclusion ： Mammography image with different algorithms of histogram enhancement, CLAHE is more outstanding than HE and HS on the low - level of noise and the enhancement of contrast.