FishEyeRecNet: A Multi-Context Collaborative Deep Network for Fisheye Image Rectification

Images captured by fisheye lenses violate the pinhole camera assumption and suffer from distortions. Rectification of fisheye images is therefore a crucial preprocessing step for many computer vision applications. In this paper, we propose an end-to-end multi-context collaborative deep network for removing distortions from single fisheye images. In contrast to conventional approaches, which focus on extracting hand-crafted features from input images, our method learns high-level semantics and low-level appearance features simultaneously to estimate the distortion parameters. To facilitate training, we construct a synthesized dataset that covers various scenes and distortion parameter settings. Experiments on both synthesized and real-world datasets show that the proposed model significantly outperforms current state of the art methods. Our code and synthesized dataset will be made publicly available.

[1]  Michel Antunes,et al.  Unsupervised Intrinsic Calibration from a Single Frame Using a "Plumb-Line" Approach , 2013, 2013 IEEE International Conference on Computer Vision.

[2]  Ming-Hsuan Yang,et al.  Deep Image Harmonization , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Bolei Zhou,et al.  Scene Parsing through ADE20K Dataset , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Jake K. Aggarwal,et al.  Depth estimation using stereo fish-eye lenses , 1994, Proceedings of 1st International Conference on Image Processing.

[5]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[6]  Kenneth Turkowski,et al.  Creating image-based VR using a self-calibrating fisheye lens , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[7]  Edward Jones,et al.  Equidistant Fish-Eye Calibration and Rectification by Vanishing Point Extraction , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Juho Kannala,et al.  A generic camera calibration method for fish-eye lenses , 2004, ICPR 2004.

[9]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[10]  Bolei Zhou,et al.  Semantic Understanding of Scenes Through the ADE20K Dataset , 2016, International Journal of Computer Vision.

[11]  Sergiu Nedevschi,et al.  Omnidirectional stereo vision using fisheye lenses , 2014, 2014 IEEE 10th International Conference on Intelligent Computer Communication and Processing (ICCP).

[12]  Stefan K. Gehrig Large-Field-of-View Stereo for Automotive Applications , .

[13]  Xianghua Ying,et al.  Can We Consider Central Catadioptric Cameras and Fisheye Cameras within a Unified Imaging Model , 2004, ECCV.

[14]  Matthew N. Dailey,et al.  Automatic Radial Distortion Estimation from a Single Image , 2012, Journal of Mathematical Imaging and Vision.

[15]  Rynson W. H. Lau,et al.  DeshadowNet: A Multi-context Embedding Deep Network for Shadow Removal , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[16]  Edward Jones,et al.  Wide-angle camera technology for automotive applications: a review , 2009 .

[17]  Yoram Singer,et al.  Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..

[18]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[19]  Xianming Liu,et al.  When Image Denoising Meets High-Level Vision Tasks: A Deep Learning Approach , 2017, IJCAI.

[20]  Deb Roy,et al.  An immersive system for browsing and visualizing surveillance video , 2010, ACM Multimedia.

[21]  Yi Zhang,et al.  Line-based Multi-Label Energy Optimization for fisheye image rectification and calibration , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Trevor Darrell,et al.  Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[23]  Hongbin Zha,et al.  Radial lens distortion correction using cascaded one-parameter division model , 2015, 2015 IEEE International Conference on Image Processing (ICIP).

[24]  Edward Rosten,et al.  Camera distortion self-calibration using the plumb-line constraint and minimal Hough entropy , 2008, Machine Vision and Applications.

[25]  Jie Sun,et al.  Calibration and Correction for Omnidirectional Image with a Fisheye Lens , 2008, 2008 Fourth International Conference on Natural Computation.

[26]  Xianghua Ying,et al.  Radial Lens Distortion Correction Using Convolutional Neural Networks Trained with Synthesized Images , 2016, ACCV.

[27]  Kiyoshi Kiyokawa,et al.  Fisheye vision: peripheral spatial compression for improved field of view in head mounted displays , 2014, SUI.