Awesome Image Rectification

A curated list of image rectification (distortion correction) and camera calibration papers.

Table of Contents

• Deep Learning Based Methods
• Traditional Methods
• Datasets
• Material
• Related Papers on Camera Calibration
• Related Libraries

Deep Learning Based

★ means that it is recommended for reading.

|Paper|Supervised|Link|Conference|Notes| |---|---|---|---|---| |Blind Geometric Distortion Correction on Images Through Deep Learning ★|Yes|Link|CVPR '19|Parameterized distortions (barrel, pincushion, and etc)| |ESIR: End-to-end Scene Text Recognition via Iterative Image Rectification||Link|CVPR '19|Scene text distortion| |Arbitrary Shape Scene Text Detection with Adaptive Text Region Representation||Link|CVPR '19|Scene text recognition| |Towards Robust Curve Text Detection with Conditional Spatial Expansion||Link|CVPR '19|Scene text recognition| |Learning to Calibrate Straight Lines for Fisheye Image Rectification ★||Link|CVPR '19|Fish eye distortion, straight line calibration| |Learning Structure-And-Motion-Aware Rolling Shutter Correction|Yes|Link|CVPR '19|Rolling shutter distortion| |Symmetry-constrained Rectification Network for Scene Text Recognition||Link|ICCV '19|Scene text distortion| |DR-GAN: Automatic Radial Distortion Rectification Using Conditional GAN in Real-Time|Yes|Link|IEEE Transactions on Circuits and Systems for Video Technology'19|Conditional GAN.Learn the mapping between distorted and clean images.No hand-crafted technique.| |FishEyeRecNet: A Multi-Context Collaborative Deep Network for Fisheye Image Rectification ★|Yes|Link|ECCV '18|Fish eye distortion, predicts camera model parameters| |MORAN: A Multi-Object Rectified Attention Network for Scene Text Recognition||Link||Scene text distortion| |GridFace: Face Rectification via Learning Local Homography Transformations||Link|ECCV '18|Face rectification| |Semantic Segmentation of Fisheye Images||Link|ECCV '18 Workshop|Sementic segmentation for fish-eye images| |Distortion-aware CNNs for Spherical Images||Link|IJCAI '18|Spherical distortion| |Fingerprint Distortion Rectification using Deep Convolutional Neural Networks||Link|ICB '18|Fingerprint distortion| |CNN-based Fisheye Image Real-Time Semantic Segmentation||Link|IEEE Intelligent Vehicles Symposium (IV) '18|Sementic segmentation on fish-eye images| |Deep View Morphing||Link|CVPR '17|3D view morphing| |Robust Scene Text Recognition with Automatic Rectification||Link|CVPR '16|Scene text distortion| |Radial Lens Distortion Correction Using Convolutional Neural Networks Trained with Synthesized Images||Link|ACCV '16|Radial lens distortion|

Traditional

|Paper|Link|Conference|Notes| |---|---|---|---| |A Robust Method for Strong Rolling Shutter Effects Correction Using Lines with Automatic Feature Selection|Link|CVPR '19|Rolling shutter distortion| |Distortion-Free Wide-Angle Portraits on Camera Phones|Link|SIGGRAPH '19|Wide-angle camera, fish eye distortion| |Restoration of Non-rigidly Distorted Underwater Images using a Combination of Compressive Sensing and Local Polynomial Image Representations|Link|ICCV '19|Underwater image distortion| |Rolling-Shutter-Aware Differential SfM and Image Rectification|Link|ICCV '17|Rolling shutter artifact correction| |From Bows to Arrows: Rolling Shutter Rectification of Urban Scenes|Link|CVPR '16|Rolling shutter artifact correction| |Line-Based Multi-Label Energy Optimization for Fisheye Image Rectification and Calibration|Link|CVPR '15|Fish eye distortion| |Radial lens distortion correction using cascaded one-parameter division model|Link|ICIP '15|Radial distortion| |Lens Distortion Rectification using Triangulation based Interpolation|Link|ISVC '15|Lens distortion| |Radial distortion correction from a single image of a planar calibration pattern using convex optimization|Link|ICIP '14|Radial distortion| |Unsupervised Intrinsic Calibration from a Single Frame Using a "Plumb-Line" Approach|Link|ICCV '13|Line detection approach| |Radially-Distorted Conjugate Translations|Link||Radial distortion| |Pixel-variant Local Homography for Fisheye Stereo Rectification Minimizing Resampling Distortion|Link||Radial Distortion| |Rectification from Radially-Distorted Scales|Link||Radial distortion| |Auto- matic lens distortion correction using one-parameter division models|Link||Radial distortion| |Automatic Radial Distortion Estimation from a Single Image|Link||Radial Distortion|

Datasets

|Paper|Paper Link|Dataset Link|Synthesized|Annotations| |---|---|---|---|---| |Parameterized Synthetic Image Data Set for Fisheye Lens|Link|Link|Yes| |Learning to Calibrate Straight Lines for Fisheye Image Rectification|Link||No|Distortion parameters and distorted lines|

Material

| Name | Course Link | Note | | ------------------------------------------------------------ | ------------------------------------------------------------ | ---------------------------------------- | | CS 131 Computer Vision: Foundations and Applications (Fall 2015-2016) | Link | Cover the basic concept of camera model. | | Review of Geometric Distortion Compensation in Fish-Eye Cameras | Link | An overview of common fish eye models. |

Papers on Camera Calibration

|Paper|Link|Conference|Note| |---|---|---|---| |Deep Single Image Camera Calibration with Radial Distortion|Link|CVPR '19|Reduce parameters from 6 to 4. Predict parameters by Simple CNN.| |Parameter-free Lens Distortion Calibration of Central Cameras|Link|ICCV '17|| |Unsupervised Vanishing Point Detection and Camera Calibration from a Single Manhattan Image with Radial Distortion|Link|CVPR '17||

Related Libraries

|Library|Link|Notes| |---|---|---| |Augmentor|Link|Image data augmentation by distortion|

Contributing

PR me related papers!