<abstract xmlns="http://www.w3.org/1999/xhtml"><p>With the rapid development of computer technology, image inpainting has become a research hotspot in the field of deep learning. Image inpainting belongs to the intersection of computer vision and computer graphics, and is an image processing technology between image editing and image generation. The proposal of generative adversarial network effectively improves the problems of poor image inpainting effect and large difference between the inpainting image and the target image, and promotes the development of image inpainting technology. In this paper, the image inpainting is based on the generation of confrontation networks. Its network structure establishes two repair paths, namely the reconstruction path and the generation path, and the two paths correspond to two groups of networks. The encoder and generator in the network respectively complete the encoding and decoding tasks based on the residual network. The discriminator also uses the patch block discriminator on the basis of the residual network to discriminate the authenticity of the image. This paper uses Places2 data set to verify the algorithm, and uses PSNR and SSIM two objective evaluation methods to evaluate the quality of the repaired image. Experiments show that the algorithm inpainting effect is better.</p></abstract>

With the rapid development of computer technology, image inpainting has become a research hotspot in the field of deep learning. Image inpainting belongs to the intersection of computer vision and computer graphics, and is an image processing technology between image editing and image generation. The proposal of generative adversarial network effectively improves the problems of poor image inpainting effect and large difference between the inpainting image and the target image, and promotes the development of image inpainting technology. In this paper, the image inpainting is based on the generation of confrontation networks. Its network structure establishes two repair paths, namely the reconstruction path and the generation path, and the two paths correspond to two groups of networks. The encoder and generator in the network respectively complete the encoding and decoding tasks based on the residual network. The discriminator also uses the patch block discriminator on the basis of the residual network to discriminate the authenticity of the image. This paper uses Places2 data set to verify the algorithm, and uses PSNR and SSIM two objective evaluation methods to evaluate the quality of the repaired image. Experiments show that the algorithm inpainting effect is better.

Image Inpainting Research Based on Deep Learning

International Journal of Advanced Network, Monitoring and Controls

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

{"article-title":"Image Inpainting Research Based on Deep Learning"}

With the rapid development of computer technology, image inpainting has become a research hotspot in the field of deep learning. Image inpainting belongs to...

Image Inpainting Research Based on Deep Learning

Data publikacji: 13 lip 2020

Zakres stron: 23 - 30

DOI: https://doi.org/10.21307/ijanmc-2020-013

Słowa kluczowe
Image Inpainting, Generation Adversarial Networks, Residual Network, Patch

© 2020 Zhao Li et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Image Inpainting Research Based on Deep Learning

Data publikacji: 13 lip 2020

Zakres stron: 23 - 30

DOI: https://doi.org/10.21307/ijanmc-2020-013

Słowa kluczoweImage Inpainting, Generation Adversarial Networks, Residual Network, Patch

© 2020 Zhao Li et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Słowa kluczowe
Image Inpainting, Generation Adversarial Networks, Residual Network, Patch