Lab notes · Video analytics
Making a tank disappear: removing moving objects from real video
Removing a moving object from real video usually fails for one simple reason: there is no clean background to put back.
The classic “invisible cloak” demo hides a person under a blue cloth. You capture one empty frame, find the blue pixels, and paste the background over them. Real footage is harder. A camouflaged vehicle is the same color as the tree line behind it, the camera pans to follow the action, and no frame in the clip is ever empty.
We tested our cloaking pipeline on a WWII Matilda II tank driving across the arena at Bovington. Instead of capturing a background, we rebuild one. Every frame is warped into a single, stabilised mosaic. Then, for each pixel, we take the median of the frames where the tank was not covering it.
The same building blocks — object segmentation, scene geometry and background reconstruction — are what you need for object removal, change detection and handling things that block the camera's view in real-world footage.
Treat the scene geometry and the model as one problem, and a “clean plate” can be rebuilt even when the camera never saw one.
Capability demo built by IPD Scientific. More lab notes →
Have footage like this?
Book a free 15-minute call, or email us a sample image or short clip. We'll tell you honestly whether computer vision can do the job.