Presentation 2012-02-09
Spotting Recognition of Time-Space Patterns From a Motion Image
Ryuichi OKA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes an optimal matching algorithm which performs segmentation-free matching between two time-space patterns. One of the two time-space patterns is a pixel trajectory called a reference time-space pattern with a fixed time interval and composed of a sequence of two-dimensional location parameter. The other one is called an input time-space pattern composed of an endless stream of image frame of gray level pixel. The proposed algorithm automatically detects a time-space pattern with a fixed time interval from an input time-space pattern without segmentation. The detected pattern is optimally similar to the reference pattern. The similarity value is optimally determined by the proposed algorithm based on dynamic programming. Recognition of a reference pattern is carried out at each time by detecting a minimum value of two-dimensional pattern composed of accumulated minimum distance values. Shape of moving object is detected by pixels with similar movement. Tracking is also realized by concatenating spotting results of primitive reference patterns.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) time-space pattern / DP / segmentation-free / motion image recognition / shape extraction
Paper # PRMU2011-188,SP2011-103
Date of Issue

Conference Information
Committee SP
Conference Date 2012/2/2(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 Speech (SP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Spotting Recognition of Time-Space Patterns From a Motion Image
Sub Title (in English)
Keyword(1) time-space pattern
Keyword(2) DP
Keyword(3) segmentation-free
Keyword(4) motion image recognition
Keyword(5) shape extraction
1st Author's Name Ryuichi OKA
1st Author's Affiliation School of Computer Science and Engineering, Aizu University()
Date 2012-02-09
Paper # PRMU2011-188,SP2011-103
Volume (vol) vol.111
Number (no) 431
Page pp.pp.-
#Pages 6
Date of Issue