| Computational intelligence techniques, such as genetic algorithms and neural networks, can be employed to solve the problems related to image processing. This paper presents the results of our studies performed to generate convolution masks used in image processing for quality improvement by using genetic algorithms. The concept of this paper is there exists a mask that can restore to some extent a degraded image to its original quality. For improving the image quality, the neighborhood convolution operator is used. For those cases for which standard designs of convolution operators do not work, genetic algorithms can provide a solution. Our results indicate that the application of genetic algorithms, gives approximately 95% success in improving the image quality when the intensity of the noise level in the sample image is low and approximately 60% successful, when the image noise level is high, with elitist and normal selection processes. |
