Earthquake Damage Visualization
Earthquakes are one of the most devastating natural disasters. One major earthquake may kill tens of thousands of people. When an earthquake occurs, casualties often occur in collapsed buildings. Post-earthquake rapid reconnaissance and detailed forensic inspections are essential to learn from failures and successes. These inspections often generate large amounts (tens of thousands) of visual data (image or video). The purpose of this project is to build computer tools for managing and analyzing the visual data so that better design and constructions methods can be developed to prevent casualties in future earthquakes.
Issues Involved or Addressed
This project has three major parts: (1) managing data, (2) reconstructing scenes, and (3) analyzing the data. The visual inspection by a single team can generate very large amounts of data and a major earthquake disaster area may be inspected by several teams. Thus, data management is critical - allowing researchers to associate data from multiple sources about the same collapsed buildings or neighborhood. The data should then be used to reconstruct 3-dimensional scenes so that researchers can visualize the structure being studied, understand the distribution of in a structure, and how the structure may have interacted with adjacent structures or ground features in the immediate vicinity. Then, the computer tools allow researchers to explore, extract, and analyze data to assist in understanding why some buildings collapse while the others stand.