|
An individual accepts optical stimulation of environment through physiological visual system, and memory is extracted from the past experience to produce a higher level of cognitive behavior. Hence, eye movement is a clue that provides the cognitive behavior. However, there is very little study has been done on the preferences and visual attention.
Eye tracking has been widely studied, especially on reading, user interface, visual design, and etc. Under the controlled experiments, these studies suggest that visual attention can be measure through the numbers of fixation. The higher the number of fixation, the higher the preference of an individual towards the object. According to Kaplan’s attention restoration theory, high voluntary attention might reduce the attention restoration; meanwhile, attention restoration has consistency towards landscape preference. For that reason, high voluntary attention might reduce landscape preference of an individual. Therefore, number of fixation and visual preference should be further study. Additional, the relationships among landscape preference, attention restoration, naturalness and image features need to be further study.
The purpose of this study is to examine the relationship between visual attention and landscape preference by adopting Kaplan’s preference matrix. The study was divided into three stages: (1) the distribution of visual attention (piolot study): 15 pieces of water landscape pictures were selected by 30 respondents, hotspot areas of the images was analyzed via spatial autocorrelation and content analysis. (2) The relationship between landscape preferences and visual attention (formal experiment): 60 pictures were provided with two different scales (macro and micro) to total of 33 respondents; pictures consist of different images, such as natural, mixed, and urban three types. Experiments focus on the relationship between naturalness, restoration, landscape preference, and number of visual attention. (3) The relationships among image features, landscape preference, and visual attention: the experiment was to examine the relationship between image features (colors, shape and texture) and landscape preference score.
At the first stage, findings suggest that Moran’s I of each image was significant (p<0.01). In other words, results indicate that respondents were in the non-random way of browsing. LISA regional verification also suggest that respondent’s eye movement will be differ due to the different image content. Findings at the second stage show that the higher the fixation number, the lower the preference score, in which it is consistent with Kaplan’s theory. Preference matrix only affected by complexity and mystery. The increment of complexity will positively reflect on the visual attention; in contrast, landscape restoration and preference will be reduced. On the other hand, the relationship between mystery and total fixation number is antagonised, the higher the mystery degree, the lower the number of total fixation. Meanwhile, the higher the mystery, the higher the preference is. At the final stage, respondents have a higher intention to green color. The tendency between image features and content of image, an individual’s preference may be affected by the composition of nature, and further research is required.
|