Strategy for the estimation of the scattering and absorption coefficients in one-dimensional participating media
In this work a strategy for the estimation of absorption and scattering coefficients in one-dimensional participating media is presented. Media are considered with the absorption coefficient in the range [0.1 to 1.0] and the scattering coefficient between [0.1-1.0]. The direct problem was solved wit...
| Authors: | , , , |
|---|---|
| Format: | article |
| Status: | Published version |
| Publication Date: | 2014 |
| Country: | Perú |
| Institution: | Universidad Nacional Mayor de San Marcos |
| Repository: | Revistas - Universidad Nacional Mayor de San Marcos |
| Language: | Spanish |
| OAI Identifier: | oai:revistasinvestigacion.unmsm.edu.pe:article/8674 |
| Online Access: | https://revistasinvestigacion.unmsm.edu.pe/index.php/fisica/article/view/8674 |
| Access Level: | Open access |
| Keyword: | inverse problem Bregman distance heat transfer entropy Havdra - Charvát. problema inverso distancia de Bregman transferencia de calor entropía de Havdra- Charvát. |
| Summary: | In this work a strategy for the estimation of absorption and scattering coefficients in one-dimensional participating media is presented. Media are considered with the absorption coefficient in the range [0.1 to 1.0] and the scattering coefficient between [0.1-1.0]. The direct problem was solved with the discrete ordinates and finite difference methods. In order to solve the inverse problem the following strategy consists of (a) find the absorption coefficient considering the scattering coefficient with an approximate value. 0.01, (b) find the scattering coefficient value using the absorption coefficient estimated in (a). The error function is defined as the difference between the measured value by the detector and the calculated by the direct problem. The algorithm used for the solution is to minimize the Bregman distance subject to the error function. Bregman distance was constructed with a related function to the entropy of Havdra-Charvát. Cases random noise tests to 2% in the measured data are presented. In order to find the best estimate we adopt as a criterion for comparison of the relative standard quadratic error. |
|---|