Presentation 2013-06-11
Criterion for image stitching based on the intensity distribution and entropy
Kenta MATSUI, Kazuaki KONDO, Takahiro KOIZUMI, Yuichi NAKAMURA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a criteria of image stitching to acquire larger panoramic images from first person view videos. Since these often have registration failures and large motion blurs caused by shaking of a wearer's head, a well stitched image is not always generated from a randomly selected image sequence. We modify the intensity model for Bayesian super-resolution to deal with the both of consistency of multiple observations for each pixel in the scene and degree of expansion of an estimated area. The proposed criteria is given as entropy of posterior in the estimation model. Its capability has been confirmed through the experiments with registration error and motion blur.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) First person view video / Panoramic images / Bayesian super-resolution
Paper # PRMU2013-32
Date of Issue

Conference Information
Committee PRMU
Conference Date 2013/6/3(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Criterion for image stitching based on the intensity distribution and entropy
Sub Title (in English)
Keyword(1) First person view video
Keyword(2) Panoramic images
Keyword(3) Bayesian super-resolution
1st Author's Name Kenta MATSUI
1st Author's Affiliation Kyoto University()
2nd Author's Name Kazuaki KONDO
2nd Author's Affiliation Kyoto University
3rd Author's Name Takahiro KOIZUMI
3rd Author's Affiliation Kyoto University
4th Author's Name Yuichi NAKAMURA
4th Author's Affiliation Kyoto University
Date 2013-06-11
Paper # PRMU2013-32
Volume (vol) vol.113
Number (no) 75
Page pp.pp.-
#Pages 6
Date of Issue