Image enhancement of chest radiography using wavelet analysis

Image enhancement of the blurred area masked by high contrast image on X-ray chest radiography, is very important for physician's diagnosis. We study the possibility of image enhancement of the masked area in spite of the size and the position of the masked area, using wavelet analysis, automatically. The image signals are decomposed to wavelet representation which lies between the spatial and the Fourier domain. The wavelet coefficient can be modified locally referring to the density of blurred image on both domains. As it is possible to keep information referring to the coordinate on the image, using features of localization of the base functions, we could transform inversively the decomposed data through modification of the coefficient. As a result, we could recognize more clearly properties of region of mediastium and lung, respectively.