Fusion of Images with Different Exposure Levels

Abstract

Most of the multi-exposure image fusion (MEF) techniques based on weighted pixel algorithms do not preserve image details, texture and vivid colors and thus lack quality. Some require pre and post processing techniques for refinement of weights. Moreover, very little work exists in block-wise/ patch-wise based MEF techniques. This research proposes a patch-wise based MEF scheme, which produces the fused image by combining the weighted average of signal structure, mean intensity, and maximum value of signal strength. Signal structure is processed by using Gray level Co-occurrence Matrix (GLCM). Incorporation of GLCM based signal structure reduces the artifacts and noise in the image thus avoiding pre/post processing. Comparison of the proposed scheme with existing MEF schemes reveals that the proposed method produces better results qualitatively and quantitatively.