Velocity Color Photo Reverse Engineering




Figure 1. Baldplate Velocity Model after reverse engineering



Objective: To generate a velocity model from a picture in a pdf file.

Procedure: Download the following files Baldplate_initial_V.png, clrbar.png, vel.png, image_deFrame.m, image_deWhiteMargin.m, invMap_colorbar.m, outlier_removal.m, velocity_reverseEngineering.m,

  1. From a pdf file (or of other formats) of interest, crop out the photo of the (a) velocity model and (b) the colorbar
  2. You can be lenient while cropping in 1(a), i.e., you can include some white margin around the color-photo (see vel.png )
  3. In 1(b), do not keep the white margin and black frame on the left and right sides. Do include the white margin and black frame on the top and bottom sides (see clrbar.png )
  4. Note down the velocity range covered by the colorbar.

    5. Explanation:

    1. The pre-processing routines will remove the white margin and black coordinate frame. Since color bar is constant horizontally, eventually only a 1d column vector (of the color mapping) is to be kept. So, no need to include the boundaries on the left and right side to start with. (In terms of image-cropping, this is easier than having to position the cursor carefully in order to include the boundary.)
    2. Run the script at the bottom of (i.e., after the final "return") velocity_reverseEngineering.m. The result is shown in Figure 1.
    3. The example provided here is for Baldplate data. Customize it for any velocity model of your choice.