A histogram equalization method is proposed. Firstly

The image is divided into low and high frequency use mean filtering. Then

various classical histogram equalization methods can be selected to deal with Lowfrequency information and highfrequency information respectively

a new method of selecting appropriate adjustment factor is presented to suppress noise from different extent. Finally

the low and high frequency information is combined to reconstruct the whole image. Experiments show that the dynamic range of the image gray levels is stretched effectively with improving contrast

image detail is well retained

noise amplification and overenhancement are also avoided.