Our group of four students completed an exciting project for our 3D Vision course. Our aim was to create a system allowing us to extract 3D coordinates with stereo vision using only one camera and a mirror. The advantage of our setup over a system of two cameras are summarized in three points. First, no camera synchronization is needed for a video data acquisition. Second, one single calibration is performed. Lastly, the system parameters (lens, sensor, spectral response, pixel size) are shared for both of the stereo images. The zoom of the camera (internal calibration) and the position of the mirror are computed by our software.
We validated our project using both a simulation we created (an artificially generated 3D scene) and a prototype we built ourselves. This allowed us to see the full potential of our system and make necessary adjustments for optimal results. Overall, our project was a challenging but rewarding experience and we were able to create a system from scratch.