目的设计一种快速有效的鼻窦 CT图像灰度归一化处理方法，消除或减小 采用不同窗参数设置的鼻窦CT显示图像之 间同质组织不同灰度上的差异。方法 数 字化图像预处理后，像素灰度做相对光密 度转换得到归一化图像；实验同时评估了 转换前后图像和病变组织的灰度准确性和 一致性。结果 鼻窦CT图像灰度归一化处 理后同一性质的病变组织灰度均值和方差 比处理前更趋于一致。结论 该方法运算 速度快，受外界条件的影响较小，有利于 后续图像识别和检索等操作。

Objective To design a rapid and effective grey scale normalization method on sinus CT images, which is used to eliminate or reduce the inconsistency of sinus CT image grey scale result from different settings of imaging factors. Methods After preliminary digitalized image processing, relative optic density transformation for grey scale are applied. We get grey scale normalized images and analyze the grey scale of before and after transformation images and pathological changed tissue regions for assessment of accuracy and consistency. Results After relative optic transformation, the mean and variance value of same pathological changed tissue grey scale are tend to more consistency. Conclusion The method has high operational speed and little outside influence, is helpful to followup image operation such as image recognition and image retrieval.