Skip to Navigation Skip to Page Content

Flying by facial recognition

Josh Lackner (left) and Michael Newman hold the quadcopter that uses facial tracking data to move the vehicle.

An increasing availability of off-the-shelf technology has given aero-hobbyists and enthusiasts access to new aircraft devices including quadcopters — model helicopters that are lifted and propelled by four rotors. Four University of Missouri College of Engineering computer science students working on a capstone project developed facial recognition software to go with the copter hardware allowing them to control a camera mounted on the aircraft.

“We made our own prototype pan/tilt axis mount using OpenCV and were able to develop facial tracking software to control it,” said Michael Newman.

“And we used the facial tracking data to move the prototype,” added teammate Josh Lackner.

The pair explained that software running on the laptop, working with the computer’s built-in web cam, was used to determine the direction they wanted the camera affixed to the bottom of the quadcopter on a pan/tilt mount to face.

“Software on the microcontroller itself interprets the wireless signal to create the movements we want,” said Newman.

Capstone adviser for the project Matt Dickinson, the department’s system administrator, said he pitched broad ideas to the capstone class and was not only pleased that a group had selected the quadcopter challenge, but that he also is pleased with the results.

“I provide the toys and the overall guidance, but they do the work. Projects last two semesters and the quadcopter project represents one that wasn’t too simple, but wasn’t so complex they couldn’t complete it,” Dickinson said. “It’s a cool project and I’m proud of what they’ve done.”

Lackner said he’s always been interested in the hardware side of things though, admitted Newman, as computer science students, the software was the more comfortable part of the project.

“Open CV was brand new to us and getting that installed took some time. We had the software part of the project done a month before our final presentation,” Newman said.

“But the first time we flew it was two weeks ago,” Lackner added.

Though their project doesn’t have a journalism component, the capstone group, which also includes Andrew Felling and Desi Elemu, was invited to present their project at the MU Reynold’s Journalism Institute’s RJInnovation showcase. Primarily, J-school student and computer science student collaborative projects are profiled at the event.

“They chose it because it was exciting,” Newman said. “It was completely different.”

Dickinson said the capstone projects are the implementation of the things the students have learned and serve to showcase the knowledge and skills they’ve acquired at MU.

Lackner and Newman both have landed great jobs as software engineers, the former headed to Garmin in Kansas City and the latter bound for Seattle with a position at Microsoft. Both men said that the quadcopter capstone project was discussed at length during their job interviews.

Newman said he has been looking at quadcopter kits to build one of his own. “I’m looking at how much it would cost me to put one together.”

“It gets pricey quickly,” Lackner said.

Considering that he will soon join the income-earning workforce Newman said, “Now I have no excuse.”