A Comparison Study: Image Restoration Based on Two Heuristic Algorithms

Creative Commons License

Dincer N., Dincer K., Arslan E.

Journal of Investigations on Engineering & Technology, vol.6, no.1, pp.28-35, 2023 (Peer-Reviewed Journal)


In computer science, optimization can be defined as maximizing or minimizing the result. Heuristic algorithms have been developed inspired by nature and to solve different optimization problems. In this study, Artificial Bee Colony (ABC) Algorithm and Firefly Algorithm (FA) have been explained in detail and a comparison between these two algorithms has been implemented. The comparison between these two algorithms is made for image restoration by using a dataset. Image restoration is the process of reducing or eliminating data loss or deterioration that may occur during the creation of an image. The loss of efficiency in the image (reducing the visual appearance of the image) is caused by noise. It is the process of obtaining the original image from the distorted image, given the knowledge of distorting factors. There are many methods applied in the literature for image restoration. In this study, two of the evolutionary algorithms have been used for this purpose and analyzed. The data set used in the study was taken from the Kaggle website. The comparison metrics are PSNR (Peak Signal-to-Noise Ratio) and MSE (Mean Squared Error). This study shows that ABC Algorithm has better results than FA on the selected 20 images dataset used for blurred image restoration. According to the results obtained, it was seen that the ABC algorithm performed % 85 better than FA.