Authors: K.K. Anisha; M. Wilscy
Addresses: Department of Computer Science, Amrita School of Engineering, Amrita VishwaVidyapeetham, Kasavanahalli, Carmelaram PO, Bengaluru 560035, Karnataka, India ' Department of Computer Science, University of Kerala, Kariavattom, Thiruvananthapuram 695581, Kerala, India
Abstract: Noise removal from colour images is an important pre-processing step in any task involving analysis of these images. Many methods have been proposed for noise removal, but they are either inefficient or quite complicated. This paper presents a simple method for removing impulse noise from colour images, where a set of standard filters are successively applied on the noisy image. The type and the order of application of these filters are determined using fuzzy genetic algorithm. The results of simulations performed on a set of standard test images for a wide range of noise corruption levels shows that the proposed method outperforms the standard procedures both visually and in terms of objective quality measures such as Peak Signal-to-Noise Ratio (PSNR), Image Quality Index (IQI) and Mean Absolute Error (MAE).
Keywords: FGAs; fuzzy genetic algorithms; FRB; fuzzy rule base; image filters; impulse noise removal; colour images; pre-processing; image analysis; simulation; noise corruption; peak SNR; signal to noise ratio; PSNR; image quality index; IQI; mean absolute error; MAE; fuzzy logic.
International Journal of Signal and Imaging Systems Engineering, 2015 Vol.8 No.4, pp.250 - 259
Available online: 10 Jul 2015Full-text access for editors Access for subscribers Purchase this article Comment on this article