Back to Search
Start Over
Geometric reconstruction of buried heat sources from a surface thermogram
- Source :
- IEEE transactions on pattern analysis and machine intelligence. 7(5)
- Publication Year :
- 2011
-
Abstract
- Attempts to reconstruct the spatial location, size, and form of buried heat sources from the measured pattern of thermograms are, in general, prohibited by the lack of a priori information about the thermal (flow) model and the source structure. In this paper, a method is introduced based on geometric reconstruction of a buried heat source configuration. This configuration must contain point sources and/or sharp edges and be confined to a plane region parallel to the surface. The medium, in which the heat source is embedded, is assumed to be homogeneous, isotropic, and of large size compared to the size of the source and to its depth below the surface. The heat flux from the surface to the ambient is assumed to follow the Newtonian cooling law. The spatial density distribution of the flux can be described by a Green function with coefficients determined by the depth of the source plane. It is possible to approximate a corresponding inverse mapping algorithm (reconstruction filter) for each source plane depth, with only one (depth) scaling parameter. The density distribution of the source structure is optimally deblurred when the reconstruction filter's scaling parameter matches the actual depth of the source plane below the surface. In the reconstruction procedure, this reconstruction filter is consecutively applied for several values of the scaling parameter. The so-called ``deblurring quality'' of the point or edge information is utilized to decide which scaling parameter achieves the sharpest image. This procedure resembles the focusing of a lens.
- Subjects :
- Surface (mathematics)
Mathematical optimization
business.industry
Plane (geometry)
Applied Mathematics
Geometry
Iterative reconstruction
Inverse problem
Reconstruction filter
Computational Theory and Mathematics
Heat flux
Artificial Intelligence
Computer Vision and Pattern Recognition
Artificial intelligence
business
Scaling
Software
Surface reconstruction
Mathematics
Subjects
Details
- ISSN :
- 01628828
- Volume :
- 7
- Issue :
- 5
- Database :
- OpenAIRE
- Journal :
- IEEE transactions on pattern analysis and machine intelligence
- Accession number :
- edsair.doi.dedup.....2e7bd437a157d2ca0620ed2605e94121