Presentation 2019-09-04
Inter-frame interpolation by height constraint for sequential human silhouette images
Daisuke Imoto, Kenji Kurosawa, Masakatsu Honma, Ryo Yokota, Manato Hirabayashi, Yoshinori Hawai,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Boundary information is important for defining the shape of an object or a person as a first approximation, and widely used for quantitative analysis in fields such as fluid, cells, medical images, and human video analysis. For example, gait-based human identification technology, which describes the shape of a person as a silhouette (boundary information), has been beginning to be used in forensic investigation. However, under low frame-rate condition, accuracy has dropped due to lack of the in-between information of a pair of successive silhouettes. In order to compensate for such an information loss, a reliable boundary tracking technique for interpolating in-between silhouettes (i.e. boundary tracking) is necessary, but a method suitable for human video analysis has not been proposed yet. In order to solve this problem, in this paper, we proposed a novel simply-extending boundary tracking method based on the assumption that the positions of the body points do not merely change along the height or vertical direction. As a result, by using the proposed method, it is elucidated that we can perform frame interpolation with much higher accuracy than the previous methods. Furthermore, the proposed method has improved the accuracy of gait-based human identification under low frame-rate conditions. It is expected that the proposed method may be effective in other applications or fields such as action recognition, physical motion science, and sports video analysis, etc.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Boundary tracking / inter-frame interpolation / human video analysis / gait-based human identification
Paper # PRMU2019-19,MI2019-38
Date of Issue 2019-08-28 (PRMU, MI)

Conference Information
Committee PRMU / MI / IPSJ-CVIM
Conference Date 2019/9/4(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Yoichi Sato(Univ. of Tokyo) / Yoshiki Kawata(Tokushima Univ.)
Vice Chair Toru Tamaki(Hiroshima Univ.) / Akisato Kimura(NTT) / Takayuki Kitasaka(Aichi Inst. of Tech.) / Hidekata Hontani(Nagoya Inst. of Tech.)
Secretary Toru Tamaki(NTT) / Akisato Kimura(OMRON SINICX) / Takayuki Kitasaka(Yamaguchi Univ.) / Hidekata Hontani(Univ. of Hyogo)
Assistant Yusuke Uchida(DeNA) / Takayoshi Yamashita(Chubu Univ.) / Hotaka Takizawa(Tsukuba Univ.) / Yoshito Otake(NAIST)

Paper Information
Registration To Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Medical Imaging / 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) Inter-frame interpolation by height constraint for sequential human silhouette images
Sub Title (in English) Application to forensic gait analysis under low frame-rate condition
Keyword(1) Boundary tracking
Keyword(2) inter-frame interpolation
Keyword(3) human video analysis
Keyword(4) gait-based human identification
1st Author's Name Daisuke Imoto
1st Author's Affiliation National Research Institute of Police Science(NRIPS)
2nd Author's Name Kenji Kurosawa
2nd Author's Affiliation National Research Institute of Police Science(NRIPS)
3rd Author's Name Masakatsu Honma
3rd Author's Affiliation National Research Institute of Police Science(NRIPS)
4th Author's Name Ryo Yokota
4th Author's Affiliation National Research Institute of Police Science(NRIPS)
5th Author's Name Manato Hirabayashi
5th Author's Affiliation National Research Institute of Police Science(NRIPS)
6th Author's Name Yoshinori Hawai
6th Author's Affiliation National Research Institute of Police Science(NRIPS)
Date 2019-09-04
Paper # PRMU2019-19,MI2019-38
Volume (vol) vol.119
Number (no) PRMU-192,MI-193
Page pp.pp.43-48(PRMU), pp.43-48(MI),
#Pages 6
Date of Issue 2019-08-28 (PRMU, MI)