Artificial Bee Colony Algorithm For Image Restoration

With the development of science and technology,a variety of digital information is transmitted in different forms in daily life.Among them,digital image information is also more widely used.However,the quality of image is impaired by various reasons,which poses a significant negative impact on thereafter image processing.Therefore,many experts and scholars committed to the study of image restoration technology,and the use of new intelligent population optimization algorithm for image restoration becomes more and more popular.In 2005,the Artificial Bee Colony(ABC)algorithm was proposed,which simulated the foraging behavior of natural honey bees.The algorithm is constructed to describe the process of searching for nectar sources with high quality,which has the advantages of few parameters,fast computation speed,strong robustness and avoiding local optimization.It is an efficient algorithm to find the global optimal solution.In this paper,we utilize the ABC algorithm for image restoration.This paper proposes a method of image denoising and blind image restoration based on ABC algorithm.Firstly,this paper proposed a method of image denoising based on ABC algorithm.By means of labeling the noise matrix,the ABC algorithm is used to estimate the value of the possible noise pixels according to the correlation between adjacent pixels of the image.Replace the original value with the evaluation value,then the image denoising is realized.We also improved the method of marking the noise matrix,and proposed a new method which combined the adjacent pixel value comparison and the four-direction method.The new method takes the edge information into consideration,reduces the probability of false noise,and uses the method of threshold judgment to control the identification of the noise pixels.The experimental result shows that the proposed method can carry out noise marking more effectively compared with the classical threshold judgment method.It can achieve excellent results in noise recognition.Comparing with other image denoising methods,we find that the proposed method can achieve a better denoising effect,both the Peak Signal to Noise Ratio(PSNR)and the Improved Signal to Noise Ratio(ISNR)of the de-noised images are higher than those produced using other methods.Secondly,the ABC algorithm is applied to blind image restoration.Under the premise of no prior knowledge of degradation process,we use the ABC algorithm to estimate the point spread function of the degradation process,then utilize the constrained least squares filtering technique to restore the image,and finally,iterates to obtain an estimation of original image.Experimental results show that the proposed method can quickly and effectively estimate the point spread function and achieve image restoration.The method makes the convergence faster and the quality of restored image better when selecting the appropriate parameters.Compared with other blind image restoration methods,such as blind deconvolution,Particle Swarm Optimization(PSO)and genetic algorithm(GA),the value of PSNR,structural similarity index(SSIM),ISNR and Visual Information Fidelity(VIF)of the restored image can reach the maximum among all the methods.This attests to the validity of the proposed method,the closest results with the original images can be obtained by using the proposed method.