3D mesh rendering of ball after taking 2D photos.
We were passionate about disrupting consumer experiences in the retail space and wanted to explore the ways in which this disruption could happen using Augmented Reality.
What it does
Let's the user take multiple pictures of their desired object. Converts (Stitches) this image into a 3D model. Ensures that the user enters a code to get access to the placement of the object and allows the user to place this object in Augmented Reality. User can place up to 15 objects of their liking.
How we built it
The frontend uses a 6 digit code to access a respective 3D image (Android Studio - Java). Placement of objects was done through(Google AR Core). Server-side code to retrieve images, authenticate users and have more CRUD features (Node JS, Express). 3D Meshing of the Object (Reality Capture by Autodesk).
Challenges we ran into
- Autodesk Reality Capture API documentation/tutorial inconsistency and errors
- Frequent transfer of images from mobile to desktop while testing different objects
Accomplishments that we're proud of
Creating realistic mesh of 3D object using solely 2D photographs, and integrating multiple pieces of software to create a large project.
What we learned
How to persist with the technical difficulties that come with integrating different components into a project
What's next for StitchAR
Scaling it into the retail space and allow a diverse range of objects to be used in the application