Multiple layers block overlapped histogram equalization for local content emphasis

Although local histogram equalization (LHE) can be used to emphasize local content in an image, several problems are associated with it. First, LHE tends to amplify speckle noise. Second, this method is not able to preserve the shape of the input histogram, which means the lost of information. Previous literatures also do not give any suggestion on selecting a proper window size for LHE. Therefore, an extension to LHE, which we call as multiple layers block overlapped histogram equalization (MLBOHE), has been introduced. This method consists of three stages, which are enhancement stage, noise reduction stage, and merging stage. Experimental results show that MLBOHE is better than some well-known histogram equalization based local enhancement methods.