Method and apparatus for dictionary learning and for reconstructing images using a dictionary

Abstract

We describe a method of providing a dictionary by sparse dictionary learning of N-dimensional image data, wherein N is a natural number greater than or equal to 3 and the method is implemented by a computer comprising a processor and a memory. The method comprises defining an N-dimensional patch for the N-dimensional image data; extracting Nbatch data sub-regions from the N-dimensional image data, using the defined N-dimensional patch; and providing the dictionary by dictionary learning using the extracted Nbatch data sub-regions. We also describe a method of reconstructing images from sparse N-dimensional image data using such a dictionary.

Dependencies are inferred from the claim text.