SR-DeblurUGAN: An End-to-End Super-Resolution and Deblurring Model with High Performance
Open Access
- 27 June 2022
- Vol. 6 (7), 162
- https://doi.org/10.3390/drones6070162
Abstract
In this paper, we consider the difference in the abstraction level of features extracted by different perceptual layers and use a weighted perceptual loss-based generative adversarial network to deblur the UAV images, which removes the blur and restores the texture details of the images well. The perceptual loss is used as an objective evaluation index for training process monitoring and model selection, which eliminates the need for extensive manual comparison of the deblurring effect and facilitates model selection. The UNet jump connection structure facilitates the transfer of features across layers in the network, reduces the learning difficulty of the generator, and improves the stability of adversarial training.Keywords
This publication has 17 references indexed in Scilit:
- Water Hyacinth (Eichhornia crassipes) Detection Using Coarse and High Resolution Multispectral DataDrones, 2022
- Drone Technology for Monitoring Protected Areas in Remote and Fragile EnvironmentsDrones, 2022
- Bayes-Probabilistic-Based Fusion Method for Image InpaintingInternational Journal of Pattern Recognition and Artificial Intelligence, 2022
- UAV Photogrammetry and GIS Interpretations of Extended Archaeological Contexts: The Case of Tacuil in the Calchaquí Area (Argentina)Drones, 2022
- Accuracy Assessment of a UAV Direct Georeferencing Method and Impact of the Configuration of Ground Control PointsDrones, 2022
- Drone Magnetometry in Mining Research. An Application in the Study of Triassic Cu–Co–Ni Mineralizations in the Estancias Mountain Range, Almería (Spain)Drones, 2021
- Generative Image Inpainting with Dilated Deformable ConvolutionJournal of Circuits, Systems and Computers, 2021
- Generative adversarial networksCommunications of the ACM, 2020
- Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent SpacePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2017
- U-Net: Convolutional Networks for Biomedical Image SegmentationPublished by Springer Science and Business Media LLC ,2015