Application Of Clustering Algorithm In Image Restoration

Image restoration technology is an important field of image processing,it mainly uses existing information in an image to estimate the unknown region,so that the image restored is visually close to the original image.Image restoration technology is widely used in real life.In recent years,image restoration based on sparse representation has become a.new research direction,in this paper,we do further research on image restoration using dictionary learning and clustering technology based on sparse representation theory.The main content of this paper is:Firstly,we introduced the background,research significance and status of image restoration at home and abroad,described the basic theory of sparse representation,some popular sparse representation algorithms about image sparse representation and K-SVD dictionary learning algorithm.Secondly,we analyzed the advantages and disadvantages of current image restoration algorithm,then,introduced the image restoration problem based on sparse representation and dictionary learning in detail,we know that the quality of the result has a great relationship with the learned dictionary.To ensure the dictionary is rich and adaptive in content,we make a cluster of training images.In addition,we described fuzzy clustering algorithm as well as its improved algorithm,s,we improved fuzzy clustering algorithm using canopy operator,Canopy-FCM algorithm can be applied in distributed platform,and compared to FCM,its the accuracy rate is improved,the inter cluster distance and the time are reduced.Finally,the simulation results in mean square error and peak signal-to-noise ratio value show that the proposed image restoration method is better than using the existing information in an image to estimate for dictionary learning.The algorithm takes good use of the similarity between in images.For the image containing rich structural information,the results are better,our algorithm make up some shortcomings of the based on K-SVD algorithm image restoration,to a certain extent,improve the quality of image restoration.