Image Inpainting Using Nonlocal Texture Matching and Nonlinear Filtering
| Autor: | Ding Ding, Jeffrey J. Rodriguez, Sundaresh Ram |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Computer science
business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Inpainting Pattern recognition 02 engineering and technology Filter (signal processing) Iterative reconstruction Similarity measure Computer Graphics and Computer-Aided Design Computer Science::Graphics Kernel (image processing) Computer Science::Computer Vision and Pattern Recognition Computer Science::Multimedia 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business Software Smoothing ComputingMethodologies_COMPUTERGRAPHICS Texture synthesis |
| Zdroj: | IEEE Transactions on Image Processing. 28:1705-1719 |
| ISSN: | 1941-0042 1057-7149 |
| DOI: | 10.1109/tip.2018.2880681 |
| Popis: | Nonlocal texture similarity and local intensity smoothness are both essential for solving most image inpainting problems. In this paper, we propose a novel image inpainting algorithm that is capable of reproducing the underlying textural details using a nonlocal texture measure and also smoothing pixel intensity seamlessly in order to achieve natural-looking inpainted images. For matching texture, we propose a Gaussian-weighted nonlocal texture similarity measure to obtain multiple candidate patches for each target patch. To compute the pixel intensity, we apply the $\alpha $ -trimmed mean filter to the candidate patches to inpaint the target patch pixel-by-pixel. The proposed algorithm is compared with four current image inpainting algorithms under different scenarios, including object removal, texture synthesis, and error concealment. Experimental results show that the proposed algorithm outperforms the existing algorithms when inpainting large missing regions in images with texture and geometric structures. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|