Research On Image Inpainting Algorithm Based On Multi-scale Transformation

As the main visual information carrier,image has been running through people's life since ancient times.Most early,the image is recorded on paper,film,paintings,etc.,which would inevitably suffer from air erosion,man-made destruction,etc.as time goes by,causing the loss of local image information.With the continuous development of science and technology,digital media technology has emerged.Image digitization has also become a trend in order to facilitate the preservation and transmission of images.However,in the process of acquiring,processing,transmitting,compressing,and decompressing digital images,image information will also lost or damaged.In order to meet the visual effects of human vision,it is extremely important to repair the missing information in the image.Based on the restoration of damaged images with different degrees,two kinds of digital image inpainting algorithms based on multi-scale transformation are proposed.The wavelet transform technology has been widely used in the field of digital image inpainting,however,the image inpainting based on wavelet transform will appear the phenomenon of edge fuzzy and not connection,which becomes a difficult problem.Based on the multi-scale and multi-directional decomposition and the traditional method of image inpainting,a new algorithm of image inpainting based on dual-tree complex wavelet transform was proposed.Firstly,the image was decomposed into low frequency and high frequency parts by using the dual-tree complex wavelet transform.Then the parts of different frequency after image decomposition were inpainted respectively.The high frequency components of the image were inpainted by the total variation model,and an improved curvature-driven-diffusion is used to repair the low frequency components.Finally,the final image is obtained by dual-tree complex wavelet transform reconstruction process.The experimental results show that the proposed algorithm is very good for the promotion of the dual-tree complex wavelet transform in image inpainting application and get better repair both in the part of texture and the part of structure.The core idea of the block-based inpainting algorithm is to find the best matching block to fill the defect block.Aiming at the problem of long time of traditional block matching process,this thesis proposes an image inpainting algorithm based on image pyramid and Coherency Sensitive Hashing(CSH).Firstly,the algorithm uses multi-scale decomposition of broken images through an image pyramid.Then,using CSH technology to repair damaged images.In detail,using the CSH to hash index the broken image A and its duplicate image B,using multiple hash functions to map A and B into a hash table,replacing the best matching block from A in the hash table A corresponding position,repeated iterations,to get the final repair map.In the end,we can obtain the result after several iterations.The experimental results show that the proposed method not only achieves good repair results,but also greatly reduces the repair time.