3D Visibility Graph Analysis
A few preliminary graphics from my thesis, an application of three-dimensional visibility graph analysis to “weak heritage” sites.
A simple 3d model is represented as an array of tiles(30×30 in the example) with each of the tiles as a vertex vi of a graph G, with an edge (vi,vj) in G if vi and vj are mutually visible from each other.
I then run standard graph centrality analysis on the graph, as shown below. The color scale is normalized for each graph, with red being the highest value, green the midway and blue the lowest.
The white “T” shaped object in the middle represents an inaccessible visual barrier.
The analysis is done in a Python program (from a simple elevation grid exported from 3ds Max via a maxscript), using the excellent NetworkX package. The results are read back into 3ds Max to generate the graphics shown.
The Clustering Centrality analysis seems particularly promising, as it can be construed as a measure of how Local or Global the visibility relationships are for any given point, which should allow for an understanding of how space is understood by both locals and visitors.
Betweenness Centrality
Closesness Centrality
Clustering Coefficient
Degree Centrality
August 27th, 2007 at 9:37 am
[…] My next step is to correlate this data with the 3D Visibility Graph Analysis. […]