Construcción de mallas triangulares no estructuradas aplicado al ajuste de superficies de objetos tridimensionales
Palabras clave:
Mallas Triangulares, Superficies 3D, Imágenes de Rango (es)Triangular Meshes, 3D surfaces, Range Images (en)
Descargas
En este trabajo se presenta una metodología para generar mallas triangulares no optimizadas de modelo no estructurado a partir de imágenes de rango. Las imágenes de rango suelen ser muy densas y por ello, reconstruir superficies con la totalidad de los puntos acarrea altos costos computacionales; para esto, se describe un modelo de selección de puntos bajo dos criterios diferentes: el primero, la medida de la distancia de cada punto a una aproximación lineal de la superficie; los puntos seleccionados en esta fase serán aquellos cuya distancia estimada sea menor o igual a un umbral establecido; y seguidamente, una selección de un conjunto de puntos, basada en la estimación de la curvatura; esta última, servirá como base para estimar la probabilidad de selección, que permitirá tener mayor densidad de puntos en regiones con altas variaciones de curvatura y menor densidad de puntos en regiones con bajas variaciones de ésta. De esta manera, es posible seleccionar un conjunto de puntos representativos de la superficie a reconstruir. Finalmente, se presentan resultados experimentales con imágenes de rango reales en formato flynn’s.
The range images usually are very dense and for that reason to reconstruct surfaces with all of the points is expensive. This work, propose a methodology to create triangular meshes not optimized of not structured models from range images, which consists in a strategy of points selection based in two criteria’s: first, the measure of the distance of each point to a lineal approach of the surface; the points selected in this phase will be those whose distance is smaller or similar to an threshold; second, a selection of points, based on the curvature estimate; this last one, will serve like base to estimate the selection probability that will allow to have bigger density of points in regions with high curvature variations and smaller density of points in regions with low curvature variations. This way, it is possible to select a set of representative points of the surface to reconstruct. Finally, experimental results are presented with real range images in format flynn’s.
Referencias
Allgower, E. y Schmidt, P. (1985), ‘An algorithm for piecewise linear approximation of an implicitly defined manifold’, SIAM Journal of Numerical Analysis 22, 322-316.
Campbell, R. y Flynn, P. (2001), ‘A survey of free-form objects representation and recognition techniques’. Comput. Vis. Image Underst. 81(2), 166-210.
Canavos, G. (1987), Probabilidad y Estadística aplicaciones y métodos, Mc-Graw Hill.
Curless, B. y Levoy, M. (1996), A volumetric method for building complex models from range images, in ‘Proceedings of SIGGRAPH ’96’, pp. 303-312.
Edelsbrunner, H. y Shall, N. (1992), Incremental topological flipping works for regular triangulations, in ‘Proc. 8th ACM Symp. Comp. Geometry', pp. 43-52.
García, M. (1997), Efficient approximation of range images through data-dependent adaptive triangulations, in ‘IEEE Int. Conf. on Computer Vision and Pattern Recognition’, San Juan, Puerto Rico.
García, M.. Domingo, S. y Basanez, L. (1997), Efficient approximation of range images through data-dependent adaptive triangulations, in ‘IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’97)’, p. 628.
Hoffman, R. (1987), ‘Segmentation and classification of range images’, IEEE Transactions on Pattern Analysis and Machine Intelligence 9, 608-620.
Hoppe, H., DeRose, T., Duchamp, T., McDonald, J. y Stuetzle, W. (1992), Surface reconstruction from unorganized points, in ‘Computer Graphics (SIGGRAPH ’92 Proceedings)', Vol. 26. pp. 71-78.
Lorensen, W. y Cline, H. (1987), ‘Marching cubes: A high resolution 3d surface construction algorithm’, Computer Graphics 21(4), 163-169.
Rieger, B., Timmermans. F., van Vliet, L. y Verbeek, P. (2002), Curvature estimation of surfnees in 3D grey-value images, in ‘16th International Conference on Pattern Recognition (ICPR'02)', Vol. 1, p. 10684.
Shewchuk, J. (1999), Lectures notes on nelatmay mesh generation, Technical report, Department of Electrical Engineering and Compurter Science, University of California at Herkeley.
Soderkvist, I. (1999), Introductory overview of surface reconstruction methods, Technical report, Department of Mathematics, Lulea University of Technology, Lulea, Sweden. Technical Report 1999-10.
Wyvill, C., McPheeters, G. y Wyvill, B. (1986), ‘Data structures for soft objects’, The Visual Computer 2(1), 227- 231.