Abstracting data from images of bicycle lanes to create a blueprint for the safest bicycle infrastructure.
In September I spent a lot of time in Amsterdam.
One of the first things I did was buy a bicycle (and a good lock).
With my helmet mounted GoPro Max I took to the streets to explore the city.
After a week spent riding a little over 75km and speaking to other cyclists (pretty much everyone!) it became clear there were many improvements that could be made.
Of course, it’s all relative. Amsterdam boasts the best cycling infrastructure for a major city that I’ve seen, but there’s always room for improvement.
And many of the people I spoke to were keen to play there part. So we decided to create a challenge around pedal power, 360 cameras and computer vision that anyone can take part in.
We’re going to abstract data from every image uploaded and use computer vision to identify areas of improvements.
Poor line markings, broken pavements, confusing signing are all things we’ll be trying to detect, alerting the city infrastructure teams when we find them.
Users can also take part in a corresponding labeling challenge, where they can earn points for labeling common hazards not detected automatically. Every label will help improve our detection models and in turn the ability to identify areas of improvements.
But it doesn’t just stop at Amsterdam.
We want YOU to help us capture images of your city, town, village, and everything in-between by bicycle. Create a Cycling Challenge on Map the Paths and get your community involved in capturing photos.
What are you waiting for? ON YOUR BIKE! Here’s some inspiration for you.
All you need to get started is a bike, mobile phone, and mount to hold it. View our recommended mobile phone setup (and action/360 camera alternatives) here.