Creating an orthomosaic from vehicle footage?

I have a video taken by a poor quality 640x480 surveillance camera on a moving vehicle. Is it possible to manually do this project, if I already know the focal length and distortion properties of the camera? I would like to manually identify points to stitch together the images by hand, and have Pix4D make the orthomosaic.

Thank you.

PIX4Dmapper can process video frames. So, yes, this is possible. I would run the video through Mapper and if the quality is not good then you could adjust the camera parameters.