Presentation 2017-09-15
Proposal of a method to detect damaged lane markings based on label propagation
Yasunari Kishimoto, Yasuhiro Iida, Sotetsu Iwamura, Makoto Kawano, Takuro Yonezawa, Jin Nakazawa, Hideyuki Tokuda,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Road marking inspection has been carried out by visual observation from the vehicle. However, visual inspection of large-scale road network by management staffs is high-cost. Therefore, in this study, we propose a method of detecting damage from the road images in low-cost. In our study, we picked up lane markings as a road marking and collects a number of images that shoot the road surfaces. Then, we dealt with those road images as a graph consisted of image as a node and similarity between images as an edge. In a few nodes, we put labels which denote “damaged” or “not damaged.” Finally we execute label propagation, all labels of nodes are estimated.Thus we got the estimation results with the accuracy rate over 85% from only 1.8% of already known correct labels.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) road marking / lane marking / damage detection / graph construction / label propagation
Paper # PRMU2017-51,IBISML2017-23
Date of Issue 2017-09-08 (PRMU, IBISML)

Conference Information
Committee PRMU / IBISML / IPSJ-CVIM
Conference Date 2017/9/15(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Shinichi Sato(NII) / Kenji Fukumizu(ISM)
Vice Chair Hironobu Fujiyoshi(Chubu Univ.) / Yoshihisa Ijiri(Omron) / Masashi Sugiyama(Univ. of Tokyo)
Secretary Hironobu Fujiyoshi(AIST) / Yoshihisa Ijiri(NAIST) / Masashi Sugiyama(Kyoto Univ.) / (Univ. of Tokyo)
Assistant Masato Ishii(NEC) / Yusuke Sugano(Osaka Univ.) / Ichiro Takeuchi(Nagoya Inst. of Tech.) / Toshihiro Kamishima(AIST)

Paper Information
Registration To Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Infomation-Based Induction Sciences and Machine Learning / Special Interest Group on Computer Vision and Image Media
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Proposal of a method to detect damaged lane markings based on label propagation
Sub Title (in English)
Keyword(1) road marking
Keyword(2) lane marking
Keyword(3) damage detection
Keyword(4) graph construction
Keyword(5) label propagation
1st Author's Name Yasunari Kishimoto
1st Author's Affiliation Nippon Telegraph and Telephone Corporation(NTT)
2nd Author's Name Yasuhiro Iida
2nd Author's Affiliation Nippon Telegraph and Telephone Corporation(NTT)
3rd Author's Name Sotetsu Iwamura
3rd Author's Affiliation Nippon Telegraph and Telephone Corporation(NTT)
4th Author's Name Makoto Kawano
4th Author's Affiliation Keio University(Keio Univ.)
5th Author's Name Takuro Yonezawa
5th Author's Affiliation Keio University(Keio Univ.)
6th Author's Name Jin Nakazawa
6th Author's Affiliation Keio University(Keio Univ.)
7th Author's Name Hideyuki Tokuda
7th Author's Affiliation Keio University(Keio Univ.)
Date 2017-09-15
Paper # PRMU2017-51,IBISML2017-23
Volume (vol) vol.117
Number (no) PRMU-210,IBISML-211
Page pp.pp.97-102(PRMU), pp.97-102(IBISML),
#Pages 6
Date of Issue 2017-09-08 (PRMU, IBISML)