Hybrid Tone Mapping with Structural and Edge-preserving Priors

  • D.Venkat Reddy, V. V. Satyanarayana Tallapragada*, K. Raghu, M. Venkat Naresh

Abstract

High dynamic range images precisely define a real-world scene in a raw file format. The content of this raw file may be modified to fit a different display unit. The process of converting the raw file into a standard dynamic range image is of two kinds. The first is global mapping, where the mapping function is global, and the same formula applies to all the pixels of the image. The second is a local mapping where different mapping functions or employing a criterion on different parts of the image. The visual quality of local mapping functions is better than that of global mapping. Still, there exist various halo effects in global mapping. In this paper, these effects are minimized using a decomposition-based global mapping. First, decompose the image into base and detail layers. -norm optimization is used on the base layer while -norm on the detail layer. An edge preserve filter and structural prior are imposed on the detail layer and base layer respectively. This kind of hybrid use of edge-preserving and structural priors remove the halo effects while preserving useful edges. The use of hybrid decomposition and corresponding optimization resulted in a high-quality tone mapped image. Simulation results prove that the proposed scheme produces a Tone mapped image index over 0.8 for most of the test images, which is better than the existing schemes.

Published
2020-06-01
How to Cite
D.Venkat Reddy, V. V. Satyanarayana Tallapragada*, K. Raghu, M. Venkat Naresh. (2020). Hybrid Tone Mapping with Structural and Edge-preserving Priors. International Journal of Advanced Science and Technology, 29(7), 5135-5143. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/23592
Section
Articles