Longuetaud, Fleur (2005) Non-destructive detection and analysis of internal characteristics of Norway spruce (Picea abies (L.) Karst.) logs by means of computed tomography. PhD thesis Sciences Forestières et du Bois, ENGREF.
Abstract
Domputerized tomography allows a direct access to internal characteristics of scanned logs on the basis of density and moisture content variations.
The objective of this work is to assess the feasibility of an automatic detection of internal characteristics with the final aim of conducting scientific analyses. The database is constituted by CT images of 24 spruces (Picea abies (L.) Karst.) obtained with a medical CT scanner. Studied trees are representative of several social status and are coming from four stands located in North-Eastern France, themselves are representative of several age, density and fertility classes. The automatic processing developed are the following:
- pith detection in logs dealing with the problem of knot presence and ring eccentricity. The accuracy of the localisation was less than one mm;
- detection of the sapwood/heartwood limit in logs dealing with the problem of knot presence (main source of difficulty). The error on the diameter was 1.8mm which corresponds to a relative error of 1.3%;
- detection of the whorls location and comparison with an optical method;
- detection of individualised knots. This process allows to count knots and to locate them in a log (longitudinal position and azimuth); however, the validation of the method and extraction of branch diameter and inclination are still to be developed.
An application of this work was a variability analysis of the sapwood content in the trunk: at the within-tree level, the sapwood width was found to be constant under the living crown; at the between-tree level, a strong correlation was found with the amount of living branches.
A great number of analyses are possible from our work...
| Item Type: | PhD Thesis (PhD) |
|---|
| Thesis Supervisor: | Leban, Jean-Michel |
|---|
| Date: | October 2005 |
|---|
| Board of examiners: | Beauregard, Robert and Lindgren, Owe and Leban, Jean-Michel and Graffigne, Christine and Mothe, Frédéric and Zoulalian, André |
|---|
| Ecole Doctorale: | ED 410 SCIENCES ET INGENIERIE DES RESSOURCES PROCEDES PRODUITS ENVIRONNEMENT |
|---|
| Discipline: | Sciences Forestières et du Bois |
|---|
| Collection (Fonds): | ENGREF |
|---|
| Institution: | ENGREF |
|---|
| Subjects: | 7. Life Sciences and Engineering |
|---|
| Uncontrolled Keywords: | Tomographie, Détection automatique, Moelle, Aubier, Duramen, Noeuds, Verticilles, épicéa, Picea abies, Billons, Qualité du bois, Défauts, Analyse d'images |
|---|
Statistiques de consultation
Repository Staff Only: edit this item
