Artificial Neural Network Method Based on Convolution to Efficiently Extract the DoF Embodied in Images (영상의 내포되어 있는 피사계 심도를 효율적으로 추출하기 위한 합성곱 기반의 인공 신경망 기법)

Jong-Hyun Kim (김종현)

doi:10.9708/jksci.2021.26.03.051

Artificial Neural Network Method Based on Convolution to Efficiently Extract the DoF Embodied in Images

Journal of The Korea Society of Computer and Information
Abbr : JKSCI
2021, 26(3), pp.51-57
DOI : 10.9708/jksci.2021.26.03.051
Publisher : The Korean Society Of Computer And Information
Research Area : Engineering > Computer Science
Received : February 5, 2021
Accepted : March 22, 2021
Published : March 31, 2021

Jong-Hyun Kim ¹

¹강남대학교

Accredited

ABSTRACT

In this paper, we propose a method to find the DoF(Depth of field) that is blurred in an image by focusing and out-focusing the camera through a efficient convolutional neural network. Our approach uses the RGB channel-based cross-correlation filter to efficiently classify the DoF region from the image and build data for learning in the convolutional neural network. A data pair of the training data is established between the image and the DoF weighted map. Data used for learning uses DoF weight maps extracted by cross-correlation filters, and uses the result of applying the smoothing process to increase the convergence rate in the network learning stage. The DoF weighted image obtained as the test result stably finds the DoF region in the input image. As a result, the proposed method can be used in various places such as NPR(Non-photorealistic rendering) rendering and object detection by using the DoF area as the user's ROI(Region of interest).

KEYWORDS

Artificial neural network, Convolutional neural network, Depth of field, Image processing, Region of interest, Object detection

Citation status

* References for papers published after 2023 are currently being built.

[confproc] Gargallo, Pau / 2005 / Bayesian 3D modeling from images using multiple depth maps / 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) / IEEE 2 : 885~891

[confproc] Silberman / 2012 / Indoor segmentation and support inference from rgbd images / European conference on computer vision / Springer : 746~760

[journal] Häne / 2017 / 3D visual perception for self-driving cars using a multi-camera system:Calibration, mapping, localization, and obstacle detection / Image and Vision Computing : 14~27

[confproc] Müller, K. / 2010 / 3-D video representation using depth maps / Proceedings of the IEEE 99(4) : 643~656

[confproc] Häne / 2011 / Stereo depth map fusion for robot navigation / IEEE/RSJ International Conference on Intelligent Robots and Systems : 1618~1625

[journal] Rafiee, Gholamreza / 2013 / Region-of-interest extraction in low depth of field images using ensemble clustering and difference of Gaussian approaches / Pattern Recognition : 2685~2699

[journal] Park, Jungwoo / 2006 / Extracting focused object from low depth-of-field image sequences / Visual Communications and Image Processing : 60771~

[other] Simonyan, Karen / 2014 / Data-driven estimation of cloth simulation models / arXiv preprint arXiv:1409.1556

[journal] Hadsell, Raia / 2009 / Learning long-range vision for autonomous off-road driving / Journal of Field Robotics : 120~144

[confproc] Eigen, David / 2014 / Depth map prediction from a single image using a multi-scale deep network / Advances in neural information processing systems : 2366~2374

[confproc] Laina, Iro / 2016 / Deeper depth prediction with fully convolutional residual networks / 2016 Fourth international conference on 3D vision (3DV) / IEEE : 239~248

[confproc] Liu / 2015 / Deep convolutional neural fields for depth estimation from a single image / Proceedings of the IEEE conference on computer vision and pattern recognition : 5162~5170

[confproc] Geiger / 2012 / Are we ready for autonomous driving? the kitti vision benchmark suite / 2012 IEEE Conference on Computer Vision and Pattern Recognition : 3354~3361

[confproc] Li / 2018 / Megadepth : Learning single-view depth prediction from internet photos / Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition : 2041~2050

[confproc] Srinivasan / 2018 / Aperture supervision for monocular depth estimation / Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition : 6393~6401

[confproc] Suwajanakorn / 2015 / Depth from focus with your mobile phone / Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition : 3497~3506

[confproc] Barron / 2015 / Fast bilateral-space stereo for synthetic defocus / Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition : 4466~4474

[confproc] Ha / 2016 / High-quality depth from uncalibrated small motion clip / Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition : 5413~5421

[report] Joshi, Neel / 2014 / Micro-baseline stereo : 8~

[journal] Klose, Felix / 2015 / Sampling based scene-space video processing / ACM Transactions on Graphics (TOG) 34(4) : 67~

[journal] Haeberli, Paul / 1990 / The accumulation buffer:hardware support for high-quality rendering / ACM SIGGRAPH computer graphics 24(4) : 309~318

[journal] Lee, Sungkil / 2010 / Real-time lens blur effects and focus control / ACM Transactions on Graphics (TOG) 29(4) : 1~7

[journal] Kraus, Martin / 2007 / Depth‐of‐field rendering by pyramidal image processing / Computer graphics forum 26(3) : 645~654

[journal] Lee, Sungkil / 2009 / Real-time depth-of-field rendering using anisotropically filtered mipmap interpolation / IEEE Transactions on Visualization and Computer Graphics 15(3) : 453~464

[journal] Yang, Yang / 2016 / Virtual dslr: High quality dynamic depth-of-field synthesis on mobile platforms / Electronic Imaging (18) : 1~9

[journal] Dong, Chao / 2015 / Image super-resolution using deep convolutional networks / IEEE transactions on pattern analysis and machine intelligence : 295~307

KJCKorea
Journal Central

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65

Artificial Neural Network Method Based on Convolution to Efficiently Extract the DoF Embodied in Images

ABSTRACT

KEYWORDS

Citation status

* References for papers published after 2023 are currently being built.

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65

Artificial Neural Network Method Based on Convolution to Efficiently Extract the DoF Embodied in Images

ABSTRACT

KEYWORDS

Statistics

Tools

Issue List

Citation status

KCI Citation Counts (0)

REFERENCES (26) * References for papers published after 2023 are currently being built.

Search PDF

Citation

* References for papers published after 2023 are currently being built.