Special Issue Article
Removal of Impulse Noise Using Fuzzy Genetic Algorithm
Digital image processing plays a key role in medical diagnosis. Medical images are obtained and analyzed to determine the presence or absence of abnormalities such as tumor, which is vital in understanding the type and magnitude of a disease. Unfortunately, medical images are susceptible to impulse noise during acquisition, storage and transmission. Hence, image de-noising is a primary precursor for medical image analysis tasks. Noise removal can be done much more efficiently by a combination of image filters or a composite filter, than by a single image filter. Determining the appropriate filter combination is a difficult task. In this paper, we propose a technique that uses Fuzzy Genetic Algorithm to find the optimal composite filters for removing all types of impulse noise from medical images. Here, a Fuzzy Rule Base is used to adaptively change the crossover probability of the Genetic Algorithm used to determine the optimal composite filters. We use Genetic Algorithm (GA) to determine composite filters that remove different levels of impulse noise from an image. In this method, the GA considers a set of possible filter combinations of a particular length, selects the best combinations among them according to a fitness value assigned to each combination based on a fitness function, and applies genetic operators such as crossover and mutation on the selected combinations to create the next generation of composite filters. We expect that the results of simulations on a set of standard test images for a wide range of noise corruption levels will show that the proposed method output performs standard procedures for impulse noise removal both visually and in terms of performance measures such as PSNR, IQI and Tenengrad values.