Perception-based image editing
Diego Gutierrez
- 04 Dec 2009
- pp 1-3
TL;DR: This work presents the recent and on-going work on advanced image editing techniques, extending current tools by leveraging the limitations of the human visual system and the wealth of information available in a high dynamic range image.
read more
Abstract: Image editing and post-processing techniques have matured over the years, making it difficult (verging on impossible) to assess whether an image has been digitally enhanced or doctored somehow. However, complex manipulations still rely on skilled user input, painstakingly painting over pixels. We present our recent and on-going work on advanced image editing techniques, extending current tools by leveraging the limitations of the human visual system and the wealth of information available in a high dynamic range image. Working in perceptual space, psychophysics plays an important role in assessing the validity of the results.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Figures
Figure 5.18: Comparison with other decomposition methods.99 
Figure 7.10: Computer generated crystal figurines. Left: photon-mapped caustics. Right: caustics painted by an artist. 
Figure 7.11: The four objects used in our first psychophysical test. From left to right: skull, vertebrae, dolphin and bull. 
Figure 3.21: A result of compositing images through relighting by using the information from our Light Source-Fitting method. 
Figure 3.22: The composited objects in the figure 3.21: the soldier and one of the elephants. Left: Original background image, with spheres showing the light directions detected for the two light probes (the Venetian mask and the wooden mannequin). Middle, top: representation of the depths assigned to the objects and the two main light sources detected. Middle, bottom: the soldier as originally photographed, relit by our algorithm and extracted from the composition in the final image. Right: the same sequence of images for the elephant. 
Figure 5.23: Multi-level decomposition (SSD09). In (a) we show the input image. In (b)-(e) we can see different layers, from the finer to the coarser level of detail. It can be observed how the illumination component is mostly captured at level (e).
Citations
Computational Imaging and Displays: Capturing and displaying richer representations of the world
TL;DR: This article looks at a few examples of Computational displays, aiming to convey the power of the joint co-design of hardware and computational algorithms, also taking into account visual perception.
1
References
Dynamic range independent image quality assessment
Tunc Ozan Aydin,Rafal Mantiuk,Karol Myszkowski,Hans-Peter Seidel +3 more
- 01 Aug 2008
TL;DR: In this article, the human visual system is used to detect and classify visible changes in the image structure, and a new metric for image quality assessment is proposed based on the detection and classification of visible changes.
303
Shedding light on the weather
Srinivasa G. Narasimhan,Shree K. Nayar +1 more
- 18 Jun 2003
TL;DR: A new physics-based model for the multiple scattering of light rays as they travel from a source to an observer is developed and it is demonstrated that a camera observing a distant source can serve as a "visual weather meter".
Screen-space perceptual rendering of human skin
Jorge Jimenez,Veronica Sundstedt,Diego Gutierrez +2 more
- 06 Oct 2009
TL;DR: A novel skin shader is proposed which translates the simulation of subsurface scattering from texture space to screen-space diffusion approximation, and has as high visual realism as a previously developed physically-based shader.
Psychophysics 101: how to run perception experiments in computer graphics
James A. Ferwerda
- 11 Aug 2008
TL;DR: This course will provide an introduction to the use of psychophysical methods in computer graphics and will teach attendees how to develop experiments that can be used to advance graphics research and applications.
38
Depicting procedural caustics in single images
Diego Gutierrez,Francisco J. Serón,Jorge Lopez-Moreno,Maria P. Sanchez,Jorge Fandos,Erik Reinhard +5 more
- 01 Dec 2008
TL;DR: This work shows by means of a set of psychophysical experiments that the resulting imagery is visually plausible and on par with photon mapping, albeit without the need for hand-modeling the underlying geometry.