Project thumbnail image

College of Engineering Unit(s): 
Electrical Engineering and Computer Science

Team: 
Lake Peterson, Jonathan Koning and Daniel French

Project Description: 

In 2022, The European New Car Assessment Program will start implementing Child Presence Detection points towards a car’s overall safety rating. To help vehicle manufacturers meet this new requirement, PreAct Technologies is using continuous wave time of flight cameras to take high frame rate video of the vehicle’s interior. This will allow for higher temporal resolution in object detection. Currently, normal cameras can only take up to 60 frames per second, while continuous wave time of flight cameras using the same amount of power can get over 200 frames per second. This is important because, if the cameras can detect people, then they can be used for detecting a crash before it happens. Having a high frame rate allows for the detection and reaction to crashes sooner than that of a low frame rate camera. Continuous wave time of flight cameras differ from that of standard cameras because these cameras pick up distance natively. This allows object detection to be done during the day and night regardless of lighting conditions.

The vision of our project was to create a real-time vehicular interior surveillance system that can detect and track all occupants and objects within the cabin of the vehicle. From here vehicle manufacturers can create safety systems that react to a collision or other emergency situation before they happen. To accomplish this, our project is to find the optimal camera placement and train a neural network to identify all occupants in the vehicle no matter where they are.

We were the first group to partner up with PreAct Technologies, enabling us to lay the groundwork for such an exciting new technology. We hope that future teams will be able to improve upon our work in the aspiration of eventually building a product that can readily aid in saving lives in real world environments.