This is a picture of Hiroshima urban fabric damage after the bomb. I will try to see if there is a possibility to measure the damage. My objective is, once I have a number, a comparison could be made with another damaged city to evaluate which city has less urban fabric destruction. A report like this -among many considerations-, could help planners to take a decision about what to save or maybe designs on tabula rasa could be needed.
My first step is to enhance the contrast. I sharpen the edges to "find" the urban fabric and see the plot graphics in 2D and 3D.
The image is converted into a binary file in order to make further analysis. This is a rough urban fabric, no buildings are recognized. The street on the left is clear in the 2D plot.
For now, I think the best way to understand the damage is to count it based on particles. The software I used, Image J will enclose what is considered to form a particle. I have chosen to see ellipses, but another tools are available. I tried them, and the particles account was exactly the same. The quantity of particles is really high: 1852 for the area selected.
I prefer to try again. When I converted the file into binary, the streets were turned black, and I'm afraid the software could interpret it as part of buildings. Then, I invert the binary image. The pattern of ellipses look better now. No central areas are enclosed, that is closer to reality, as the damage is even everywhere in the picture selected. Surprisingly, the account is pretty close to the previous one: 1823
A flood could show very different results, as water is covering the streets and part of the houses. The image surface seems softer, in consequence, the damage is less, as no house is disintegrated. Let us see a flood in a Mexico neighborhood, 2007. The quantity of particles is 431. To have more accurate results in comparisons, both images should be the same scale.
Flood in Mexico, 2007