Data gathered during scientific exploration is often very difficult to read and interpret in the form that it is saved - large text files. To improve this interpretation, we can create graphs to better exemplify what the data is. An even better improvement is to have graphical programs to show changes in data over time, over distance, or any other form of measurement.

This specific scientific visualization is the mapping of underwater cisterns from Malta. The first step in interpreting this data is a SLAM algorithm. The SLAM takes in the sonar scan data and creates probability maps for wall locations surrounding the robot's location. The SLAM algorithm program generates an output text file which is the input data file for this scientific vizualization. This program creates a simulated environment from the SLAM map using the marching cubes algorithm. A .m mesh file is created in order to save the vizualization and allow for texturing and manipulation by other students.

The Marching Cubes algorithm is used to create surfaces by drawing triangles at the crossing point from "inside to outside" of the volume. It is a way to draw volumetric data - exactly the data we have from the cistern exploration. The marching cubes algorithm generates a grid division of voxels and determines a crossing point on edges of these voxels. The vertices of these crossing points are stored, and a .m mesh is generated based on the vertices and faces.

Algorithm Overview:

1. Read four slices into memory;
2. Scan slices and create cubes;
3. Calculate cube index based on 8 vertex scalars;
4. Look up edge (triangle) list from table;
5. Interpolate surface edge intersections;
6. Calculate unit normals using central differences;
7. Output triangles and vertex normals;

This project was designed for use in the ICEX program's project of mapping underwater cistens in Malta.

The following images are screen shots of a simple keyhole cistern that was mapped during the trip to Malta.