Presentation 2007-10-26
Signboard Recognition by Consistency Checking of Local Features(Internationa Session 6)
Jihoon Kim, Taik Heon Rhee, Kee-Eung Kim, Jin Hyung Kim,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The problem of recognizing signboards in street scenes is defined as matching the input image to pre-stored 2D signboard images. This problem is not as simple as it appears to be due to arbitrary drawings and relative 3D positions. We approach this problem by first matching local features of input image to those of images in the database. These matches are later verified with the global features, namely the homographic consistency and the color consistency. The well-known SIFT feature is used as the local feature and the homographic consistency checking is performed using RANSAC, a randomized method for finding an optimal match. In order to handle signboards with large perspective distortions, several templates with different perspective transformations are generated a-priori. In our experiments, the proposed method achieves up to 95% recognition rate, showing good results despite the highly distorted input images.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Signboard recognition / Homography / Perspective Transform / SIFT / RANSAC
Paper # PRMU2007-118
Date of Issue

Conference Information
Committee PRMU
Conference Date 2007/10/18(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 ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Signboard Recognition by Consistency Checking of Local Features(Internationa Session 6)
Sub Title (in English)
Keyword(1) Signboard recognition
Keyword(2) Homography
Keyword(3) Perspective Transform
Keyword(4) SIFT
Keyword(5) RANSAC
1st Author's Name Jihoon Kim
1st Author's Affiliation Computer Science Department, KAIST()
2nd Author's Name Taik Heon Rhee
2nd Author's Affiliation Computer Science Department, KAIST
3rd Author's Name Kee-Eung Kim
3rd Author's Affiliation Computer Science Department, KAIST
4th Author's Name Jin Hyung Kim
4th Author's Affiliation Computer Science Department, KAIST
Date 2007-10-26
Paper # PRMU2007-118
Volume (vol) vol.107
Number (no) 281
Page pp.pp.-
#Pages 6
Date of Issue