Urban Rating Pairwise & Elo Rating
Workflow Overview
The methods we used in this project allowed for a pairwise image comparison to obtain results where children are most likely to play in an urban area. The Elo rating method was used to make a decision concerning the urban areas of the three study areas. A normal distribution of photos was collected to populate the dataset used in this study. The dataset is populated from a python application with a Tkinter toolkit, which is a Graphical User Interface that allowed the visual representation for the pairwise comparison of playground images in North Vancouver, Stevenson and East Vancouver. The images are taken from a phone camera and stored in SQLite relational database management system. The goal of our methodology is to compare 50 chosen pairs from each study area in order to determine where children are more likely to play in an urban area.
