An Optimized Filtering Technique For The Enhancement Of Digital Mammographic Images

Image mining deals with the extraction of knowledge from image data. As the image may contain unwanted details the pre-processing plays a key role in image mining. Pre-processing is performed through filtering of image which enhances the quality of the image. Filtering can be done using spatial domain filters or frequency domain filters. This paper aims at finding the optimum frequency that yields high quality pre-processed image for the Butterworth bandpass filter, a frequency domain filter. The quality of the pre-processed filtered image is evaluated using PSNR and MSE values.