Abstract:Neural network based methods for fisheye distortion correction are effective and increasingly popular, although training network require a high amount of labeled data. In this paper, we propose an unsupervised fisheye correction network to address the aforementioned issue. During the training process, the predicted parameters are employed to correct strong distortion that exists in the fisheye image and synthesize the corresponding distortion using the original distortion-free image. Thus, the network is constrained with bidirectional loss to obtain more accurate distortion parameters. We calculate the two losses at the image level as opposed to directly minimizing the difference between the predicted and ground truth of distortion parameters. Additionally, we leverage the geometric prior that the distortion distribution depends on the geometric regions of fisheye images and the straight line should be straight in the corrected images. The network focuses more on the geometric prior regions as opposed to equally perceiving the whole image without any attention mechanisms. To generate more appealing corrected results in visual appearance, we introduce a coarse-to-fine inpainting network to fill the hole regions caused by the irreversible mapping function using distortion parameters. Each module of the proposed network is differentiable, and thus the entire framework is completely end-to-end. When compared with the previous supervised methods, our method is more flexible and shows better practical applications for distortion rectification. The experiment results demonstrate that our proposed method outperforms state-of-the-art methods on the correction performance without any labeled distortion parameters.

Image Data Loss Minimized Geometric Correction for Asymmetric Distortion Fish-eye Lens

Estimating Radial Distortion for Fish-eye Lens Based on Collinear Constraint

Linear fitting interpolation based on FOV division for correcting wide angle fish-eye lens distortion

Camera lens radial distortion correction using two-view projective invariants.

Correction of Barrel Distortion in Fisheye Lens Images Using Image-Based Estimation of Distortion Parameters

Unified Optical Distortion Correction Method for Imaging Systems Using A Concise Geometrical Transformation Model

Fish-eye image correction method and technique based on geometrical imaging model

Two distortion correcting methods for fisheye images

A Comprehensive Overview of Fish-Eye Camera Distortion Correction Methods

Distortion correction of fisheye lenses using a nonparametric imaging model

Fisheye lens distortion correction

Correction method for fisheye image based on the virtual small-field camera.

Fisheye image Distortion Correction Based on Spherical Perspective Projection Constraint

Orthorectification of Fisheye Image under Equidistant Projection Model

A Distortion Correction Method Based on Actual Camera Imaging Principles

Unsupervised Fisheye Image Correction Through Bidirectional Loss with Geometric Prior

Pixel-variant Local Homography for Fisheye Stereo Rectification Minimizing Resampling Distortion

An efficient deep learning-based framework for image distortion correction

A distortion correction algorithm for fish-eye panoramic image of master-slave camera

A Method of Close-Range Calibration of Fisheye Lens Camera

A distortion correction algorithm for fish-eye panoramic image of master-slave visual surveillance system