Style-based tone mapping for HDR images

Abstract

Figure 1: Different artists may prefer different tone reproductions of the same HDR image. The same artist may also choose to produce different styles based on the situational/contextual considerations. The images reproduced by a professional artist using a different tool (top) are replicated using our operator (bottom). Our algorithm can also learn the styles of an artist and produce results consistent with them. AbstractIn this paper we propose a different approach to high dynamic range (HDR) image tone mapping. We put away the assumption that there is a single optimal solution to tone mapping. We argue that tone mapping is inherently a personal process that is guided by the taste and preferences of the artist; different artists can produce different depictions of the same scene. However, most existing tone mapping operators (TMOs) compel the artists to produce similar renderings. Operators that give more freedom to artists require adjustment of many parameters which turns tone mapping into a laborious process. In contrast to these, we propose an algorithm which learns the taste and preferences of an artist from a small set of calibration images. Any new image is then tone mapped to convey the appearance that would be desired by the artist.

Yang Chengzong and the University of Science and Technology of China

“…The emphasis is very much on aligning exposure images. In the case of only camera motion between exposure sources, i.e., the captured scene is rigid and on a plane, simple operations such as translation [56], homography [51], correlation matching [1], and phase cross correlation [62] can be used to crack this challenge. For dynamic scenes, most methods seek non-grid dense correspondences (e.g., joint homography and SIFT [14], joint homography and optical flow [21], and RANSAC [11,16]) between images.…”

Advanced tools and framework for historical film restoration

“…Wang et al 30 transfer tone and color properties of photographs taken with high-end DSLR cameras to photographs taken with cell phone cameras with lower image quality. Another interesting style transfer method based on learning the user's preferences has been presented by Akyuz et al 31 specifically for HDR images. In our case, the style transfer occurs from the film negative to the film positive, where both the positive and negative can be chemically and physically degraded at various degrees.…”

Automated parameter tuning for tone mapping using visual saliency

“…Many of the tone mapping operators are derived from the human visual system, and try to preserve perceptual factors, such as brightness, contrast, and visibility during tone mapping. More recently, tone mapping operators tend to consider user assistance [27,11], display devices [30], and artistic freedom [1]. A thorough review of current tone mapping algorithms can be found in books by Reinhard et al [37] and Banterle et al [6].…”