Presentation | 2000/9/14 An Improvement of Aspect Identification Rates of Polyhedron by Using Correlations between x-y-Coordinates of Feature Points Hiroyasu Sakamoto, Teiji Ohta, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | As a tool of 3D-model-based computer vision, the authors have proposed "aspect identifier matrices" for convex polyhedron which can identify what combination of the object's faces, edges, etc. (aspect) is observed in an image and can solve a problem of corresponding feature points. Although the identifier matrices have been improved so far, all the identifier matrices have been constructed for each horizontal and vertical coordinates, separately. In this report, we propose new identifier matrices which combine two coordinates in each matrix. Paying attention to that nonlinear distortion(error) of feature points in camera(perspective) images introduces patterns of variances and cross-covariances which are particular for each aspects, our identifier matrices can compensate for the patterns to reduce mis-identification rate. It is shown that the compensation can work more effectively when two coordinates are handled simultaneously than when handled separately. and, consequently, the proposed identifier matrices can reduce misidentification rates more than before. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Model based vision / Identification of aspect and feature point correspondence / Nonlinear error of perspective images / Covariance of error |
Paper # | PRMU2000-73 |
Date of Issue |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2000/9/14(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) | An Improvement of Aspect Identification Rates of Polyhedron by Using Correlations between x-y-Coordinates of Feature Points |
Sub Title (in English) | |
Keyword(1) | Model based vision |
Keyword(2) | Identification of aspect and feature point correspondence |
Keyword(3) | Nonlinear error of perspective images |
Keyword(4) | Covariance of error |
1st Author's Name | Hiroyasu Sakamoto |
1st Author's Affiliation | Kyushu Univ. Design() |
2nd Author's Name | Teiji Ohta |
2nd Author's Affiliation | Sohjoh University |
Date | 2000/9/14 |
Paper # | PRMU2000-73 |
Volume (vol) | vol.100 |
Number (no) | 311 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |