In sports broadcasts, how do they display those lines on the field?

Something I’ve wondered for a long time: in various sports, there are lines superimposed on the picture in various situations - to display the line of scrimmage in football, for instance, or in throwing contests to show the distance to beat. You get the idea. I see how people can walk “over” those lines, I suppose that’s simple green screen technology, but how do those lines get there in the first place, with a moving camera? Are there invisible markers on the field that the computer system picks up on?

Not an expert but any chance, but I can see it being possible they have software that makes a graph of the field and hooks the cameras’ paths on to that. You can make each player a dot and go from there.

Real time video software is getting really quite good at being able to detect what’s part of the scenery and what is not - even without a green screen.

For a moving camera, it needs to know where it’s pointing, which can be done by a variety of different methods:
Markers in the field of view - like those suits with little dots on that people used to wear for motion capture, you can place markers (sometimes they look like 2d barcodes) around the field of view - the software can detect these in the streamed footage and work out the angle and orientation of the thing it’s looking at.

Or the camera/mount can be equipped with shaft encoders on the tilt axes, or gyroscopic sensors.

So when the system knows which way it’s looking, it can be configured to know what’s there when the field is empty - from that point on, separating the foreground and background is more or less a matter of subtraction - and because you know what the orientation of the camera is, you can substitute a computer rendered background to fit.

That’s a vast oversimplification, but it’s something like that. The technology/approach has even partly trickled down into computer webcam drivers now - on my laptop here, I can make it take a picture of the empty room, then separate out anything that’s different (i.e. me) and display it with a different background. Only works for stationary camera position obviously - as my webcam has no position sensors.

Lots of info here:

How Stuff Works article “How the First-Down Line Works”

Cool! Thanks, everyone!

Wow, a lot more sophisticated than I could have imagined. That’s very neat.

My son, the programmer, was telling me about this a while ago. I didn’t understand much, but it is extremely sophisticated. Imagine, I can even recall (from maybe 40 years ago) when the introduced instant replay, how proud they were of the latest technology.

When they first started doing it, the line was overlaid over anything green, including in some cases portions of the players’ uniforms.

For the wide-angle shots, they can probably use the same hash marks and yard lines we humans use, for this purpose. I don’t know if they actually do (there might be even better methods), but it’d certainly be an option.

I remember reading an article about the technology back when it first came out and was floored that they actually took the curve of the field into account. I’d assumed the whole thing was a lot simpler than that.

Last I knew, this still happened some times, for parts of uniforms within the range of grass color on that particular field. Much less often than at first, though. Probably gets rarer every year.

“Yes, over here, I noticed in last week’s game there was a line projected on the football field! Please explain how it was done!”
“Uh, yeah, well, whenever you notice something like that… a computer did it.”
“Yes, alright, yes, but there was also-”
“Computer.”

C’mon guys. Maybe you need a refresher. It’s all ball bearings these days. Or software.
In my studio at work, I use a Tricaster similar to this one. It has software that lets me put any number of overlays and underlays. Mangetout is overthinking things a good bit. It’s all software done in the switcher, not hardware.

You can apply the layer - in this case the line of scrimmage or down line - to 1 or more camera inputs, and toggle it off and on. When you go to camera 1 with Bob on the sideline, it’s not there, but switch to camera 2 showing the field from high above the 50, it will be there. Camera 3 is of the uprights, so no line, but cam 4 can show it, and so on.

It’s all in the switcher, independent of cameras, angles, or any hardware. It’s actually a pretty fun machine to play with.

We used to do a website with Fran picking fantasy football teams every week. I’d shoot Fran and Chris discussing playerslike this here. That chalkboard behind them could just as easily be a video of an LA freeway chase, or a fishing show. You can insert screens with input from a dozen sources to make every show unique. Here’s the program intro I made mostly with the Tricaster. He got tired of doing the show, so we’ve quit for a while. We may start it up next season, but I doubt it.

He still does a lot of appearances on CNN and Fox these days, and the picture you see is from our studio in our (Fran’s) office. I’ll put a background up with a green screen, and feed it to Fox. They split it, and put it up beside whoever is on.

For those of you who like to learn about how things work, watch this introto the Tricaster. There are other brands of switchers of course, but they generally all have the same capabilities.

Whether it is an 18-wheeler in the parking lot of a sporting event, or a megachurch broadcasting on Sunday morning, a switcher is the machine - and software - that makes pulling all the camera and graphic inputs possible.

Well, sure, the software puts the overlay on the broadcast, but the real question was how does it know where to put it?

What you’re describing wouldn’t work if the camera is moving. The line has to be moved in real time to compensate for the movement of the camera so that it always appears to be in the same position on the field. So the system needs to have information on how the camera is positioned and oriented at all times. The link in engineer_comp_geek’s post explains it pretty well.

It must be quite sophisticated. IIRC, when it was first used the developers were able to charge $20k per game when it was used. I don’t know if there is now competition or if the pricing has changed but it had to have been quite a lucrative undertaking for those that developed it.

Sports Illustrated just did a story on the birth of the yellow line last month.

Sportsvision (the guys behind the yellow line) offered it to Fox the first year for $25,000 per game. FOX turned them down, CBS and ABC wanted to think on it, ESPN took them up on the deal for the first season.

Sportsvision comes in three days before the game to run cable and take a laser survey of the field. They take swatches of grass before and during the game to make sure they have the right colors. Then the game broadcast is delayed 2/3 of a second to allow the first down line to be inserted.

Lots of technology just so we can holler about the spot of the football. Heck, I find it tough to go to a game in real life now. I keep looking for the yellow first down line on the field.

Here’s a nice video showing how it works (same as the HowStuffWorks story above, but with more visuals and animations).

Agreed. For the television viewer it is probably the most significant advancement since the instant replay. You get hooked on it very quickly.

It seems to me that since the introduction of it there are a lot fewer incidents where they delay the game to “bring out the chains”. I wonder if the refs are asking the booth what the line showed before they call for a measurement.