Have a personal or library account? Click to login
oflibnumpy & oflibpytorch: Optical Flow Handling and Manipulation in Python Cover

oflibnumpy & oflibpytorch: Optical Flow Handling and Manipulation in Python

Open Access
|Nov 2021

Figures & Tables

jors-9-380-g1.png
Figure 1

Flow fields in the two possible frames of reference: “source” means all pixels at time t1 are mapped to a new location at time t2, while “target” means all pixels at time t2 are matched with a different previous location at the time t1.

jors-9-380-g2.png
Figure 2

Schematic of the flow field composition operation. To obtain the correct coordinate mapping xt1xt3 from Flows 1 and 2, it is necessary to select the flow vector from 23 which starts at the end position of the tracked feature at time t2, or xt2. Therefore, it is not sufficient to simply add the vectors of the two flow fields: this would be equivalent to selecting the flow vector from 23 which starts at xt1 (assuming the “source” frame of reference is used).

DOI: https://doi.org/10.5334/jors.380 | Journal eISSN: 2049-9647
Language: English
Submitted on: Jun 21, 2021
|
Accepted on: Nov 10, 2021
|
Published on: Nov 26, 2021
Published by: Ubiquity Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2021 Claudio S. Ravasio, Lyndon Da Cruz, Christos Bergeles, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.