Inspiration
It was great to solve a problem that actually exists in real life and that will have a positive impact on the working environment of people in the building management industry. The challenge provided a great opportunity for the team to dive deep into OpenCV and augmented reality, two topics that all of us had wanted to learn more about for a long time. By splitting up the different tasks among the team members and then educating each other on how they were implemented we were able to cover a lot of ground and learn about many different techniques.
What it does
Point your camera at a building equipment diagram and the system will detect all the nodes and edges and will convert it into a digital graph. The graph is then drawn in augmented reality on top of the paper diagram.
How I built it
First we deskew the image by detecting parallel lines and then transforming it. In the deskewed image we then detect all nodes and edges and classify extracted symbols. The symbols are overlaid on the original input image in augmented reality. The application is written in Python with OpenCV.
Challenges I ran into
Making the detection of nodes and edges robust under different lighting conditions required a lot of iteration and researching different techniques.
Accomplishments that I'm proud of
There are many great individual parts in the system, such as the deskewing algorithm which was implemented without any framework.
What I learned
Diving deep into OpenCV and uncovering its many different functionalities was a great learning experience.
What's next for GrARph the Building
The next step is to make the system more performant so it can work at a higher framerate.
Log in or sign up for Devpost to join the conversation.