Answer to Question #10798 in MatLAB | Mathematica | MathCAD | Maple for faizan
GChoose your favorite images, and add noise to them, either scratches, as described above, or salt and
pepper noise. Adapt the damage detector so that you can choose whether to turn it on when the discon-
tinuity is detected only on one layer, on all three layers, or on some average computed on all the three
layers, and restore the image. Start with the algorithm applied only on horizontal slopes, as described in
this paper, and improve it adding also a check on vertical slopes.
1. Next, modify the threshold, and plot the error between the original image and the restored image,
as a function of the threshold. Which of the given norms seems to give results coherent with what
your own eyes can see?
2. Does the optimal threshold depend on the amount of salt and pepper noise that you add to the
3. Do you get better results if you diagnose a damaged pixel using information from all neighboring
4. What is the color of nothing? I mean: you compute the error by subtracting the original and the
Unfortunately, your question requires a lot of work and cannot be done for free. Submit it with all requirements as an assignment to our control panel and we'll assist you.