image-registration

What are steps used in a intensity-based registration

What are the steps in intensity-based registration?

If I have a floating image matching to a stack of images (of which only one is the best match), what are the steps in registration?

Why is medical image normally use intensity-based for registration?

Solution

  • Image registration is an optimization problem, therefore we need to apply the algorithm iteratively (repeat many times until finding the optimal solution)

    1. Similarity measure.
    2. Optimization
    3. Rendering
    4. Repeat 2+3 steps on floating image until we found the optimal solution (based on minimizing or maximizing the cost function) or exceed the iteration limit up.

    Why is medical image normally use intensity-based for registration?

    Because of it more accurate than feature-based registration.