Tone Mapping for Single-shot HDR Imaging Johannes Herwig, Matthias Sobczyk and Josef Pauli Intelligent Systems Group, University of Duisburg-Essen, Bismarckstr. 90, 47057 Duisburg, Germany Keywords: High Dynamic Range Imaging, Tone-reproduction Operators, Noise Reduction, Image Segmentation. Abstract: The problem of tone mapping for HDR (high dynamic range) to LDR (low dynamic range) conversion is introduced by a unified framework considering all the usual processing steps. Then the specific problem of single-shot HDR is outlined where special emphasis is taken on the effect of the greater noise floor of those images when compared to the usual exposure bracketing approach to HDR. We herein tailor the popular tone mapping operators proposed by Reinhard for single-shot HDR. A region-based approach for preprocessing any HDR image in order to increase SNR and perceptual sharpness is introduced as an extension to our initial tone mapping framework. The results are compared with respect to specially developed baseline tone mappers and an extensive subjective evaluation is performed. 1 INTRODUCTION Tone mapping operators are used in a high dynamic range (HDR) image acquisition and processing chain (Reinhard et al., 2010) as the final completion. Tone mapping allows displaying or printing a HDR image on LDR (low dynamic range) media by compressing the wide tonal range of the HDR image into an im- age with lower tonal sampling. Thereby the bit-depth of the HDR image pixel commonly is 32-bit floating point but LDR images are only 8-bit unsigned inte- gers. Tonal compression should be able to preserve the overall contrast, textural details and color fidelity of the original image (Frazor and Geisler, 2006). Often only a HDR image is capable of captur- ing the real dynamic range of any natural scene, but both consumer digital displays and printing technolo- gies are only capable of dealing with low dynamic range images (DiCarlo and Wandell, 2000). The crux thereby is that photographers usually want to create a ”true” reflection of their visual experiences which they want to convey to their viewers, but due to lim- ited capturing and displaying technologies any photo- graph can never be as visually rich as the real scene. Therefore, the tone mapper is crucial in delivering an image that ”feels” as naturalistic as possible when viewed at low dynamic range. Although, tone mapping operators are developed with the (in-)capabilities of the human visual system (HVS) in mind, their design can be considered more an art than engineering. This is also because of the vast amount of factors that influence the sensation of an image where lots of assumptions are to be made. For example, apart from the tonal richness of the par- ticular HDR image additional properties of the view- ing conditions and the audience are to be considered: specific display technology, viewing distance, ambi- ent lighting, emotional state, cultural background, etc (Bodrogi and Khanh, 2012). In this paper, we present several extensions and enhancements of the tone mapper for photographic tone reproduction originally introduced by Reinhard (Reinhard et al., 2002). Reinhard has developed dif- ferent tone mapping operators in the past which are commonly acknowledged for their naturalistic results thereby advancing this field of research (Reinhard et al., 2010). His and other operators usually assume that the HDR image was created by fusing a series of differently exposed LDR images of the same scene. 1.1 Tone Mapping for Single-shot HDR In our application scenario only one image is taken with a pixel depth of 12-bit unsigned integer, which means there are 4096 different values that we want to tone map to 8-bit or 256 different values per color channel. We use the so-called RAW imaging mode of a digital camera which directly stores the raw but color balanced pixels from the digital imaging sensor. For our tone mapping we exploit the fact that mod- ern digital consumer cameras internally have higher analog-to-digital conversion (ADC) capabilities than 145 Herwig J., Sobczyk M. and Pauli J.. Tone Mapping for Single-shot HDR Imaging. DOI: 10.5220/0004695401450152 In Proceedings of the 9th International Conference on Computer Vision Theory and Applications (VISAPP-2014), pages 145-152 ISBN: 978-989-758-003-1 Copyright c  2014 SCITEPRESS (Science and Technology Publications, Lda.)