Generalisation of textured 3D city models using image compression and multiple representation data structure
2013 (English)In: ISPRS journal of photogrammetry and remote sensing (Print), ISSN 0924-2716, Vol. 79, 68-79 p.Article in journal (Refereed) Published
Texture is an essential part of 3D building models and it often takes up a big proportion of the data volume, thus makes dynamic visualization difficult. To compress the texture of 3D building models for the dynamic visualization in different scales, a multi-resolution texture generalization method is proposed, which contains two steps: texture image compression and texture coloring. In the first step, the texture images are compressed in both horizontal and vertical directions using wavelet transform. In the second step, TextureTreeis created to store the building color texture for the dynamic visualization from different distances. To generate TextureTree, texture images are iteratively ｓegmented by horizontal and vertical dividing zone, e.g. edge or background from edge detection, until each section is basically in the same color. Thentexture in each section is represented by their main color and the TextureTree iscreated based on the color difference between the adjacent sections. In dynamic visualization, the suitable compressed texture images or the TextureTree nodes are selected to generate the 3D scenes based on the angle and the distance between user viewpoint and the building surface. The experimental results indicate that the wavelet based image compression and proposed TextureTree can effectively represent the visual features of the textured buildings with much less data.
Place, publisher, year, edition, pages
2013. Vol. 79, 68-79 p.
Three-dimensional Building model, Texture Compression, Multiresolution Image, Multiple Representation Data Structures, Dynamic Visualization
Remote Sensing Other Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-48170DOI: 10.1016/j.isprsjprs.2013.02.008ISI: 000318889600006ScopusID: 2-s2.0-84875254960OAI: oai:DiVA.org:kth-48170DiVA: diva2:456890
QS 20120328. Updatad from Submitted to Published. 201306272011-11-162011-11-162014-06-24Bibliographically approved