Physical Security Unity Simulation with Object Recognition

This project involved building a physical pan-tilt camera in Unity to emulate a real-world camera with the same physical specifications. The objective was to create a framework that could be used to simulate a real-world camera and the environment which it would be placed in. Within this environment, the camera should be able to track targets of interest and classify them when possible. The video to the right presents the final version of this project using YOLOv5 for object recognition. An environment was built in unity which spawned objects onto a conveyor belt, and a moving wall was spawned and given a set path for repeated obfuscation of the target objects.

The video to the left was the end of the first sprint. At this point, the physical security camera had been built to spec in Unity and it was manually controlled. Raytracing was drawn from the security camera game object to target game object as it moved along predefined waypoints in the workspace.

Previous
Previous

"ExPORT" A novel VR/AR Interaction Technique