Source HDR video by Grzegorz Krawczyk, RNL demo courtesy of Paul Debevec.
Tremendous progress in the development and accessibility of high dynamic range (HDR) technology that has happened just recently results in fast proliferation of HDR synthetic image sequences and captured HDR video. When properly processed, such HDR data can lead to very convincing and realistic results even when presented on traditional low dynamic range (LDR) display devices. This requires real-time local contrast compression (tone mapping) with simultaneous modeling of important in HDR image perception effects such as glare (A), day and night vision (BC), visual acuity (D). We propose a unified model to include all those effects into a common computational framework, which enables an efficient implementation on currently available graphics hardware. We develop a post processing module which can be added as a final stage of any real-time rendering system, game engine, or digital video player, which enhances the realism and believability of displayed image streams.
This project received the 1nd Best Paper Award at the SCCG 2005 conference. The paper, slides from the conference presentation, and the demonstration video are available for download in section reference.
Visual adapation while entering a tunnel. Source HDR video by Grzegorz Krawczyk.
While displaying the sequence of HDR frames, it is important to note that the luminance conditions can significantly change from frame to frame. The human vision reacts to such changes through the temporal adaptation processes. The time course of adaptation differs depending on whether we adapt to light or to darkness, and whether we perceive mainly using rods (during the night) or cones (during the day). Therefore it is necessary to account for this process in a perceptually enabled tone mapping operator.
The same scene in photopic (left) and scotopic vision (right). Source HDR video by Grzegorz Krawczyk.
The human vision operates in three distinct adaptation conditions: scotopic, mesopic, and photopic. The photopic and mesopic vision, typical in daylight conditions, provide good color perception and details visibility. However in scotopic range (for instance moonlight illumination), where only rods are active, color discrimination is not possible and perception of spatial details becomes limited. While we observe an image on a display device our vision usually operates in the mesopic range, therefore it becomes necessary to simulate the effects typical to scotopic vision to convey the appearance of night scenes.
Glare effect. (RNL demo courtesy of Paul Debevec)
Due to the scattering of light in the optical system of the eye, sources of relatively strong light cause the decrease of contrast in their vicinity - glare. Such an effect cannot be naturally evoked while perceiving an image on a display due to different viewing conditions and limited maximum luminance of such devices. It is therefore important to account for it while tone mapping.
Source HDR video by Grzegorz Krawczyk, RNL demo courtesy of Paul Debevec.
We demonstrate our method in combination with an HDR video player. The player renders the compressed HDR video stream to a floating point texture, which is then processed by the algorithm described in the paper. The sample results of our method including the perceptual effects are shown in the following demo.
Perceptual Effects in Real-time Tone Mapping
Grzegorz Krawczyk, Karol Myszkowski, Hans-Peter Seidel,
In: Proc. of Spring Conference on Computer Graphics, 2005. 1st Best Paper Award.
[bibtex]
[paper]
[slides]
[video]