A study of nonlinear image enhancement with homomorphic filtering in wavelet domain

Abstract

In many application areas of image processing, the main topic of this paper is the enhancement of mammographic image which is one of diagnostic imaging modality. In spite of advances in image resolution and film contrast, film mammography remains one of diagnostic imaging modality where the image interpretation is very difficult. The major reason for poor visualization is the small size of mass and the minor difference in x-ray attenuation between normal and malignant glandular tissues. For the contrast enhancement of mammographic image, in this paper, we analyze the conventional method based on wavelet transform and newly propose the homomorphic filtering method, which is applied on the top of conventional wavelet framework. An unsharp masking technique is applied and performed in wavelet domain. In addition, simple nonlinear enhancement and a de-nosing stage that preserves edges using wavelet shrinkage are computed into this technique. The use of the proposed homomorphic filtering technique is described in order to enhance the contrast in the mammographic images. The use of homomorphic filtering has advantage that it inverts the exponential operation caused by the radioactive absorption and stretches the histogram of mammographic image. Experimental results show that the homomorphic filtering method improves the contrast in mammographic images such that the contrast improvement index is increased by two fold compared to the conventional wavelet-based enhancement technique.