@inproceedings{c7c995478d664616bb5097c76d0560a8,
title = "Exploring View Pattern and Analysing Pupil Size as a Measure of Restorative Qualities in Park Photos",
abstract = "Earlier evidence suggests that natural scenes are in general more restorative than built scenes but we know very little about which specific components or structures in nature promote restoration. Using eye tracking this study investigates which components e. g. flowers, benches and water, people focus on when evaluating if the space is restorative. A sample of 38 photos of small urban green spaces (pocket parks) was preselected for low and high restoration likelihood. Photos were presented on a screen and eye movements were registered for each photo, followed by a rating task about restoration likelihood. Visual analysis such as heat maps and scan paths were done as well as analysis of correlation between restoration likelihood ratings and pupil size. Results show that although the view pattern among individuals was variable, people tended to focus on similar components. The components drawing most attention were benches and other people. Regarding the pupil size we found a negative correlation with restoration likelihood, the smaller the pupil size the higher value on restoration likelihood. We interpret the result that restoration is the opposite to emotional arousal. Arousal is in other studies found to cause pupil dilation. Photos on parks promoting restoration and causing relaxation, would hence result in constriction of the pupil.",
keywords = "eye tracking, pocket parks, restoration likelihood, eye tracking, pocket parks, restoration likelihood",
author = "Helena Nordh and Caroline H{\"a}gerh{\"a}ll and Kenneth Holmqvist",
year = "2010",
language = "English",
isbn = "9789066056435",
series = "Acta Horticulturae",
publisher = "International Society for Horticultural Science",
pages = "767--772",
booktitle = "Proceedings of the IInd International Conference on Landscape and Urban Horticulture Bologna, Italy, June 9 - 13, 2009",
address = "Belgium",
}