|
Title: Incorporating Saliency Map into Appraisal of Crossing Island Facilities: From Point of View of Drivers' Visual Attention
Accession Number: 01373473
Record Type: Component
Abstract: In this study, crossing island facilities are evaluated from the conspicuity point of view. This paper provides an evaluation method to determine which kind of facilities can draw the visual attention of a driver easily with the aid of a saliency map. A visual scan experiment with the aid of eye tracking system was conducted to collect fixation percentage data. The evaluation method of crossing island facility was conducted based on image characteristics. The saliency-based model which is based on a model of visual attention built up by Itti was set up and improved through the process of Gaussian pyramid creation, visual feature extraction and conspicuity map generation as well as the saliency map generation. Mean intensity of salient zone on the saliency map is selected as an indicator to calculate fixation percentage of crossing island facility and the regression model is helpful for the evaluation of conspicuity of crossing island facility at pedestrian refuge island together with MATLAB and Photoshop.
Supplemental Notes: This paper was sponsored by TRB committee AND10 Vehicle User Characteristics
Monograph Title: Monograph Accession #: 01362476
Report/Paper Numbers: 12-2458
Language: English
Corporate Authors: Transportation Research Board 500 Fifth Street, NW Authors: Xu, ZhiGuan, HongzhiYang, Xiao KuanYan, HaiPagination: 28p
Publication Date: 2012
Conference:
Transportation Research Board 91st Annual Meeting
Location:
Washington DC, United States Media Type: Digital/other
Features: Figures
(12)
; References; Tables
TRT Terms: Uncontrolled Terms: Subject Areas: Highways; Pedestrians and Bicyclists; Safety and Human Factors; I82: Accidents and Transport Infrastructure; I83: Accidents and the Human Factor
Source Data: Transportation Research Board Annual Meeting 2012 Paper #12-2458
Files: TRIS, TRB
Created Date: Feb 8 2012 5:09PM
|