This program generates models, giving the user the choice to either convert an image into a model or create a parallel plane model.
- Generate Property Models from Images
- Generate Parallel Plane Property Models
- Creates Binary File of the Models
- Visualiation of the Models
Follow these instructions to get a copy of the project and used it at your will.
Type the following in terminal to clone the project.
$ git clone https://github.com/GISIS-UFF/CreatePropertyModels.git
Execute this command to install the required libraries necessary to run the program:
$ python3 -m pip install -r requirements.txt
To use it, configure the parameters in the parameters.txt
file.
After configuring, run the main.py
file by typing in the terminal:
$ python3 main.py
The model dimentions(Nx and Nz) are exactly the image dimentions. Example: in an image with the following dimentions
Have these parameters:
Nx = 1150 [m]
Nz = 648 [m]
Model parameters need to be the same size as the layers in your model. Example: A model with 3 layers needs to have parameters like this:
VP_Velocity = [x, y, z]
VS_Velocity = [x, y, z]
Rho_Value = [x, y, z]
- If you turn on the "VS Velocity Approximation" option, you don't need to fill the VS_Velocity array.
Parameter priority is set from top to bottom. Example: a 2-layer VP model with 2000 [m/s] and 4500 [m/s] respectively from top to bottom needs to have parameters like this:
VP_Velocity = [2000, 4500]
The same logic as explained above applies, but now you need to specify the Interfaces array. You need to fill the array with the depths (in meters) where you want to place a layer. Example: If you need a model of 441x501 and want a layer at 250 meters, then:
Interfaces = [250]
Another example: If you need a model of 1001x501 and want layers at 250 meters, 400 meters, and 500 meters, then:
Interfaces = [250, 400, 500]
- Remember, if you have n interfaces, you will have n+1 layers, so your other parameters need to be of size n+1.
Here's some examples of the code running:
Original image:
Generated model:
Original image:
Generated model:
- Davi Melonio