Geometric Calibration of Micro-Lens-Based Light-Field Cameras using Line Features



Abstract
We present a novel method for the geometric calibration of micro-lens-based light field cameras. Accurate geometric calibration is the basis of various applications. Instead of using sub-aperture images, we directly utilize raw images for calibration. We select appropriate regions in raw images and extract line features from micro-lens images in those regions. For the entire process, we formulate a new projection model of a micro-lens-based light field camera, which contains a smaller number of parameters than previous models. The model is transformed into a linear form using line features. We compute the initial solution of both the intrinsic and the extrinsic parameters by a linear computation and refine them via non-linear optimization. Experimental
results demonstrate the accuracy of the correspondences between rays and pixels in raw images, as estimated by the proposed method.


What's New!!
2016/03/10 We have updated our source code (version 2.0) and Lytro Illum dataset.
2016/03/01 Our extended paper has been accepted in TPAMI (IF=5.781)!!


Publication
Geometric Calibration of Micro-Lens-Based Light-Field Cameras using Line Features
Yunsu Bok, Hae-Gon Jeon and In So Kweon
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), Accepted

Geometric Calibration of Micro-Lens-Based Light-Field Cameras using Line Features
Yunsu Bok, Hae-Gon Jeon and In So Kweon
European Conference on Computer Vision (ECCV), Sep 2014
[PDF] [Supp] [Source Code] [Dataset(Lytro)] [Dataset(Lytro Illum)] [bibtex]
*Last update : 2016.03.09., Version 2.0 (Improving a performance of calibration, and updating dataset (Lytro Illum)) 

If you have questions, please contact ysbok@rcv.kaist.ac.kr
Comments