Image Registration Technique for Recovering Rotation, Scale and Translation

Image data obtained by sampling an object at different times, or from multiple perspectives, will be saved to different coordinate systems. In order to transform these sets of data into a single coordinate system, you must go through a process called image registration.

Image registration is necessary for integrating data obtained through multiple means and using it for various projects or comparison.

When dealing with medical imaging, specifically data obtained for the same patient at different times, elastic registration is involved in order to deal with elastic deformations of the imaged body parts.

Image Registration Technique for Recovering Rotation, Scale and Translation is a piece of source code that was developed to provide users with an accurate registration scheme that takes scale, rotation, and translation into account. Based on a hybrid approach, it should work in both the frequency domain and time domain.

Of course, you will need Matlab to use this source code. The Demo code needs to be extracted in the Matlab directory, and you just need to type “test_main” in the command window to try it out. If you are happy with the code after the performance evaluation, you can go ahead and purchase the complete source code.