An Enhanced Image Fusion Framework Using Morphological Operations Based Unsharp Masking

Abstract

The idea of infrared (I-R) and Visible (V-I) image fusion is to integrate multiple source images and to produce a single useful informative image. Nowadays the image generated from I-R and V-I image fusion process has been used majorly in surveillance and remote sensing applications. It plays a crucial role in improving visibility and situation awareness especially in surveillance applications. This paper is introducing an enhanced I-R and V-I image fusion framework. A new Enhancement method is constructed using morphological operations based unsharp masking has been used in this algorithm for enhancing I-R and V-I source images. This enhancement method has produced high quality enhanced results which in return tremendously helped in improving the final fusion result. In this algorithm curve-let transform has been used to produce “detailed” and “approximation” coefficients. Integration of “approximation” coefficients is done through using “PCA fusion rule”. Combining of “detailed” coefficients is done with using “max fusion rule”. Fused image reconstruction is done with using inverse curve-let transform. The proposed fusion framework has produced superior results and outperformed than the similar existing fusion frameworks in terms of both visual quality and metrics values in comparison.