Blind Geometric Distortion Correction on Images Through Deep Learning

Xiaoyu Li 1           Bo Zhang 1           Pedro V. Sander 1           Jing Liao 2          

1 The Hong Kong University of Science and Technology           2 City University of Hong Kong

Fig. 1. Our proposed learning-based method can blindly correct images with different types of geometric distortion (first row) providing high-quality results (second row).

Abstract

We propose the first general framework to automatically correct different types of geometric distortion in a single input image. Our proposed method employs convolutional neural networks (CNNs) trained by using a large synthetic distortion dataset to predict the displacement field between distorted images and corrected images. A model fitting method uses the CNN output to estimate the distortion parameters, achieving a more accurate prediction. The final corrected image is generated based on the predicted flow using an efficient, high-quality resampling method. Experimental results demonstrate that our algorithm outperforms traditional correction methods, and allows for interesting applications such as distortion transfer, distortion exaggeration, and co-occurring distortion correction.

Downloads

Paper [arXiv] Supp [PDF] Code [Github Repository]

BibTeX

@inproceedings{li2019blind,
    title={Blind Geometric Distortion Correction on Images Through Deep Learning},
    author={Li, Xiaoyu and Zhang, Bo and Sander, Pedro V and Liao, Jing},
    booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
    pages={4855--4864},
    year={2019}
}

Acknowledgements

We thank the anonymous reviewers for their valuable comments. This work was partly supported by CityU of Hong Kong Start-up Grant No. 7200607/CS, and Hong Kong GRF Grant No. 16208814.