Pages: | Home | Geographic and Satellite Information | Google Maps | Process and Implementation | Outcomes and Application |
Our Story:
Our team began as part of a Software Design class project at Olin college. However, our story began with one of our developers/team members hiking off-trail in Chile. After 30 days of hiking, he had no idea how far he had traveled. The only way he was able to determine the distance was by screenshotting the available thumbnails of Chilean military topographic maps, drawing the routes in with a pen, and then using a headphone cable to measure out the distances. When maps were brought up as an interesting project topic, he recalled his Chilean adventure and the challenges of determining the length of his off-path hiking trip.
Originally, our goal was to create maps from satellite images by detecting streets and roads. We started with pattern detection done by the computer. As we got further into line and feature detection, we realized that our accuracy levels were too low for what we wanted to accomplish. Which is how we realized that people are much better at recognizing patterns than computers, but computers can give statistics that people cannot. We decided to pivot with this notion in mind. Our program now allows users to write in their patterns, and have the computer return the statistics. For example, someone going on a road-trip could draw in their path of travel and get estimations of distance traveled, nearby restaurants, etc. This approach to mapping allows the users to focus on the pattern and the computer to focus on the statistics. Now, after processing the satellite images into usable pictures, we use the Google Places API to gather information about each latitude and longitude.