The Facet Approach to Optic Flow,
Abstract
The facet approach requires low level image processing techniques to be based on fitting each local image neighborhood with a function and interpreting all processing in terms of what the processing does to this locally fit function. Using the facet approach the authors develop a different meaning of the usual optic flow equation. They show that it represents the intersection line of the isocontour plane with a successive image frame. The intersection line on the successive frame contains the possible match points. A unique match point can be selected by requiring it to have the same brightness as the given pixel. It is shown that this procedure amounts to assuming that all derivatives of third or higher order are negligible and that gray tone intensity and first partial derivatives in row, column, and time must match.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1983
- Accession Number
- ADP001197
Entities
People
- Jong Soo Lee
- Robert M. Haralick
Organizations
- Virginia Tech