Abstract
Application of weight coefficients of the bilateral filter used to determine weighted similarity metrics of image ranges in optical flow computation algorithm that employs 3-dimension recursive search (3DRS) was investigated. By testing the algorithm applying images taken from the public test database Middlebury benchmark, the effectiveness of this weighted similarity metrics for solving the image processing problem was demonstrated. The necessity of matching the equation parameter values when calculating the weight coefficients aimed at taking into account image texture features was proved for reaching the higher noise resistance under the vector field construction. The adaptation technique which allows excluding manual determination of parameter values was proposed and its efficiency was demonstrated.
| Original language | English |
|---|---|
| Title of host publication | Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016: Proceedings of the International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016 |
| Subtitle of host publication | Proceedings of the International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016 |
| Publisher | American Institute of Physics Inc. |
| Volume | 1783 |
| ISBN (Electronic) | 9780735414457 |
| DOIs | |
| Publication status | Published - 10 Nov 2016 |
| Event | International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016 - Tomsk, Russian Federation Duration: 19 Sep 2016 → 23 Sep 2016 |
Conference
| Conference | International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016 |
|---|---|
| Country | Russian Federation |
| City | Tomsk |
| Period | 19.9.16 → 23.9.16 |
Fingerprint
ASJC Scopus subject areas
- Physics and Astronomy(all)
Cite this
Application of bilateral filtration with weight coefficients for similarity metric calculation in optical flow computation algorithm. / Panin, S. V.; Titkov, V. V.; Lyubutin, P. S.; Chemezov, V. O.; Eremin, A. V.
Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016: Proceedings of the International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016: Proceedings of the International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016. Vol. 1783 American Institute of Physics Inc., 2016. 020181.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Application of bilateral filtration with weight coefficients for similarity metric calculation in optical flow computation algorithm
AU - Panin, S. V.
AU - Titkov, V. V.
AU - Lyubutin, P. S.
AU - Chemezov, V. O.
AU - Eremin, A. V.
PY - 2016/11/10
Y1 - 2016/11/10
N2 - Application of weight coefficients of the bilateral filter used to determine weighted similarity metrics of image ranges in optical flow computation algorithm that employs 3-dimension recursive search (3DRS) was investigated. By testing the algorithm applying images taken from the public test database Middlebury benchmark, the effectiveness of this weighted similarity metrics for solving the image processing problem was demonstrated. The necessity of matching the equation parameter values when calculating the weight coefficients aimed at taking into account image texture features was proved for reaching the higher noise resistance under the vector field construction. The adaptation technique which allows excluding manual determination of parameter values was proposed and its efficiency was demonstrated.
AB - Application of weight coefficients of the bilateral filter used to determine weighted similarity metrics of image ranges in optical flow computation algorithm that employs 3-dimension recursive search (3DRS) was investigated. By testing the algorithm applying images taken from the public test database Middlebury benchmark, the effectiveness of this weighted similarity metrics for solving the image processing problem was demonstrated. The necessity of matching the equation parameter values when calculating the weight coefficients aimed at taking into account image texture features was proved for reaching the higher noise resistance under the vector field construction. The adaptation technique which allows excluding manual determination of parameter values was proposed and its efficiency was demonstrated.
UR - http://www.scopus.com/inward/record.url?scp=85006022086&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85006022086&partnerID=8YFLogxK
U2 - 10.1063/1.4966475
DO - 10.1063/1.4966475
M3 - Conference contribution
VL - 1783
BT - Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016: Proceedings of the International Conference on Advanced Materials with Hierarchical Structure for New Technologies and Reliable Structures 2016
PB - American Institute of Physics Inc.
ER -