Autoconfig

Inspiration

We are avid photographers and noticed that there were rich applications for smartphone cameras that are very under-powered compared to something like a DSLR. We used a single board computer as a portable platform for writing computational photography software that integrates with a DSLR.

What it does

A smart viewfinder that provides a platform for writing apps that interface with a DSLR and a machine learning assisted auto mode for automatically adjusting camera settings based on environment and subject matter.

Our current version is just able to take a picture.

How we built it

Our project consists of a Qt5 application. In the future will be running on a single board computer with an LCD module acting as a camera viewfinder. The application interface allows the user to change settings. Libgphoto2 provides the bridge between the beaglebone and the DSLR camera.

Challenges we ran into

CMake configurations are not always for the faint of heart, especially when trying to bypass Qt Creator altogether. We had a lot of back and forth discussion on what kind of data to use along with issues trying to scrape Flickr data.

Accomplishments that we're proud of

Something that can control a camera.

What we learned

lots of C++, building a Qt5 application and cross-compiling to the beaglebone, libgphoto2, and about cameras.

What's next for Autoconfig

AI assistance feature. Add more cool features like an HDR mode, focus stacking, object removal via seam carving (http://www.faculty.idc.ac.il/arik/SCWeb/imret/index.html). Building the project into a usable open source platform for computation photography and digital photography enthusiasts who know a bit of programming and want to take things to the next level.

Built With

• c++
• cmake
• libgphoto2
• qt5