Quantitative and qualitative evaluations of image enhancement techniques

Due to the growing demand to obtain information about the structure, therefore the importance of digital image processing (DIP) takes place to improve the quality of the photographic image. Several types of image enhancement techniques took place in order to improve the quality of the corrupted images. Unfortunately, these techniques suffer from the inability of keeping fine details although they have good performance when dealing with images corrupted with low percentage of noise. In this paper, a comparative study on traditional techniques in both spatial and frequency domains with self-organizing artificial neural networks (ANN) techniques for recovering images corrupted with different percentages of impulse noise 10%-90%. From the simulation results, it was clear that the performance of a self organizing artificial neural networks technique is better than the traditional techniques in both spatial and frequency domain when dealing with noise ratios 30%-40%. In addition to the ability of self-organizing ANN techniques to recover images corrupted with higher noise ratios when using higher mask sizes