Abstract
Visual preservation and reproduction of subtle and feeble scene details is challenging during conversion from Hyper-spectral Image (HSI) to Low Dynamic Range (LDR) color images due to poor scene area to pixel density ratio, where a large geographical plane is mapped to a single pixel. This results in an averaging effect or a partial volume effect that affects the visual distinction of fine subtle details. Secondly, HSI data reduction to LDR color image involves sub-band averaging which again affects details distinction, especially structures of feeble strength. The proposed approach addresses this issue and improves the details in the scene for preservation and enhancement by modifying an existing HSI to LDR color image conversion method. An interim, detail enhancement stage is introduced. The results show a clear improvement in visual content reproduction and the numerical quantification also agrees to the same. The results are compared using three data sets and two other methods including the original method being modified.