eli5 Why do security cameras with movement outlines have trouble keeping up with fast moving people/objects?

630 views

I saw a video of a dog running around under popular and the security cam had a green box following his movement and settling around him, but would be behind as he was running, wouldn’t it instantly be able to follow him since, yknow, its a machine and can think really fast?

In: Technology

3 Answers

Anonymous 0 Comments

Just because it’s a computer doesn’t mean it’s magic. Image processing is really hard for computers, and a security camera system isn’t exactly going to be using a top of the line CPU. The tracking box is drawn as fast as the processor can process the image and detect motion, whereas the video itself is more as less as real time as a digital video is going to get. The faster the object is moving, the more behind the box is going to be.

Anonymous 0 Comments

Cameras work by taking in a certain amount of light during an exposure and sampling that. If an object is moving fast enough, the light the camera receives will include parts from all of it’s motion, creating what is know as motion blur. This blur effect, well, blurs the outline of the object that software is trying to detect. The more motion blur there is, the less sure the software can be about what it is even seeing. It’s a hard problem to remove motion blur from a picture frame, as you are basically trying to recreated data that even in the picture. It’s like sharpen tool in photoshop to undo a blur effect. It doesn’t really work.

Anonymous 0 Comments

It’s possibly a software issue. Yeah, computers think really fast, but if a program’s coded like garbage it can’t utilize that properly.

Imagine if I were trying to walk down the street, but whoever was giving me directions kept looping me around blocks and winding me back and forth down streets in a zigzag.

Oh and I also have to stop and look at my map every 10 steps.

Not saying that is necessarily the problem but that sort of stuff can slow programs down.