Research On Color Image Inpainting Based On Gradient Change And Structure Tensor

Digital image inpainting refers to the process of repairing or rebuilding the missing information of the damaged image by using a specific image inpainting algorithm,so that the observer can't easily detect that the image has been repaired or damaged.At present,the technology has been widely used in many research areas,such as the restoration of old photos,the removal of image text,the preservation of cultural relics and so on.According to the basic principle of color image inpainting,this paper applies the theory of structure tensor to color image restoration,and discusses the color image restoration algorithm based on gradient change and structure tensor to improve the effect of restoration.The main research results and contributions of this paper are summarized as follows:1.The thesis summarizes image restoration technology and the research status,analyzing the main difficulties and key technologies.2.Due to the incorrect filling order and the fixed size of patch,the traditional exemplar-based image inpainting algorithm tended to cause the image structure fracture and texture error extension.A improved color image inpainting algorithm based on adaptive adjustment with gradient variation was proposed.Firstly,using isophote curvature to constrain the filling order to overcome the discontinuity of the edge structure caused by the incorrect filling order;Secondly,in order to solve the problem of the staircase effect in rich texture regions,it adaptively adjusted the sample patch size according to the variation of local gradient;Finally,using the local search method to find the best matching patch.The experimental results show that the improved algorithm can achieve good results for different types of damaged images and effectively maintain the smoothness of the edge structure of the image.Compared with the original algorithm,the power signal-to-noise ratio(PSNR)of the final result is improved by approximately 1d B~ 4d B,and the structure similarity(SSIM)has improved.3.Because the structure tensor of color image not only includes the intensity information of the local area,but also contains the main direction of the gradient of the specific pixel neighborhood,and it can fully reflect the local geometric structure of the image.Therefore,a color image inpainting algorithm based on structure tensor was proposed in this thesis.Firstly,the proposed algorithm uses the structure tensor to define the data items to ensure that the structure information of the image can be transmitted more accurately.Then,the size of sample patch is adaptively selected according to the average coherence of the structure tensor,and the constraint of thestructure tensor is added to the matching criterion to reduce the error matching rate.The experimental results show that the improved algorithm uses the characteristics of the structure tensor to realize the inpainting of the damaged color image with different structural features and has the better restoration effect to the complicated linear structure and the texture area.Compared with other algorithms,the structure similarity(SSIM)has increased,and the peak signal-to-noise ratio(PSNR)has improved by 1d B to 3d B.