Perception of colour gradients in real and computer-simulated scenes: effects on depth

Lead Research Organisation: University of Bradford
Department Name: Faculty of Life Sciences

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
 
Description This project set out to explore interactions between different visual cues, for the perception of depth and shape. In this project we have studied how humans distinguish between depth, light source, and material properties, and how varying these various parameters impacts on depth and shape perception. In one strand of the project, we studied how colour and luminance information can influence depth perception. We found that realistic colour/luminance gradients can be used to perceive shape (Lovell, Bloj & Harris, 2012), and that although the visual information is often ambiguous, observers can quickly learn the relationship between location, lighting and shape, so that the cue becomes reliable (Harding, Harris & Bloj, 2012). We explored the effects of colour variation on shape-from-shading information. Contrary to previous findings, we discovered that colour does not always enhance perceived depth, as would be expected if the visual system exploited information about the statistics of the natural environment (Clery, Bloj, Harris, 2013, Journal of Vision). Instead, the effects of colour on shape from shading are likely to be due to process in very early vision, where the separate colour and luminance process channels first interact.In the other main strand of the project, we used optimal integration models to demonstrate that shading information can be used as a reliable cue to depth, when combined with binocular disparity, and that a classic model, the Maximum Likelihood Estimator, provides an excellent description of how the brain combines the visual information (Lovell, Bloj, Harris, 2012, Journal of Vision. 12, 1, p. 1-18. doi: 10.1167/12.1.1). Related work has also shown how depth complexity within a display can alter depth perception (Harris, 2014, Zeiner, Spitschan & Harris, 2014). We have also, for the first time, explored how these two visual cues are combined over time. Our research has demonstrated that shape from shading can be a 'quick and dirty' cue to shape at very short viewing durations (where disparity is not useful). At longer durations, binocular disparity becomes the more reliable source of visual information.
Exploitation Route To make new progress in visualisation techniques, it is becoming increasingly important to account for human perception and to progress algorithmic research based on a firm empirical and theoretical basis. Our work involved revealing some of the basic perceptual mechanisms required for visual information to be combined for 3D perception. The results have implications for the development of 3D displays and for the development of animated 3D content. The work is also useful for informing those interested in appearance in architecture, museum displays, video game design and other applications requiring an understanding of the brain computations that underlie vision. With this in mind, Harris co-organised a cross-disciplinary conference for vision scientists, technologists and engineers: Predicting Perception: 3rd International Conference on Appearance, 2012.
Sectors Digital/Communication/Information Technologies (including Software),Manufacturing, including Industrial Biotechology,Culture, Heritage, Museums and Collections,Retail,Other
 
Description The work is useful for informing those interested in appearance in architecture, museum displays, video game design and other applications requiring an understanding of the brain computations that underlie vision. With this in mind, Harris co-organised a cross-disciplinary conference for vision scientists, technologists and engineers: Predicting Perception: 3rd International Conference on Appearance, 2012.
First Year Of Impact 2012
Sector Digital/Communication/Information Technologies (including Software),Education,Culture, Heritage, Museums and Collections
Impact Types Cultural,Societal
 
Description Leverhulme Project Grant
Amount £200,344 (GBP)
Organisation The Leverhulme Trust 
Sector Charity/Non Profit
Country United Kingdom
Start 03/2018 
End 03/2021