Determining of translation and rotation of an object from its Radon transform via neural network

A flexible approach to the determination of translation and rotation of an object from Radon transform data is presented. Translation and rotation of an object can be effected in Radon space. This also works in reverse, and the Radon transform of two identical objects can be compared to determine the rotation and the translation between the objects. This gives the absolute location of the test object if the location of the reference object is known. The technique for achieving this involves the use of a neural network that has been trained to determine the orientation by determining how many lines of the sinogram have to be restacked and to determine the translation by finding the shift of each line