Two Camera's

[tomy]

For this topic I am referring to this post:

So you have one camera connect to the cRio with the ip address of (192.168.90.254), when you do this you can not go onto any laptop and go to that ip address and watch what is going on. If you change the ip address to (10.xx.yy.zz) where (xx.yy) equals your team number and (zz) equals a number from 1-21(<-i think) you can go to any laptop and connect to the wireless router, open up a internet browser and type in (10.xx.yy.zz) and view the camera(<-i think).

so if you put two cameras on in theory, you should be able to see it from the driver station and from any laptop. So.... in the regionals couldn't you technically set your router to auto and still connect with it via wired and wireless and see the camera still from the stands.

i think that first should look into this. they can still configure our routers how ever they want and then tell that team what/how they can connect to the camera.

this is my thought of using two cameras.

Will it work or will first say no cause that person then have a driver station and be controlling the robot?

[Greg McKaskle]

You are onto something ... , but then you are off.

It is possible to plug the camera(s) into the dlink and give them IP addresses in the 10.te.am.11+ range. I'd suggest making them static rather than dynamic IP assignment.

This is super useful from the programming laptop since you can now use vision assistant or a web browser to connect to the camera, make changes, capture images, monitor, etc.

This is also useful from a dashboard. Rather than ask the cRIO to forward all of the traffic from one port to the other, the camera can do this with no cRIO effort.

In the shop, when your dlink is in access point mode, the dlink publishes a network with an SSID name. If another laptop, iphone, etc knows the name of the network, any security settings, and the IP of the camera, they can log into the camera.

At a competition, the dlink is reconfigured to be a bridge -- basically the same as the black bridge from last year. Since it is a bridge, it doesn't publish a network, it extends one. Additionally, the network SSID that FIRST creates for each team is encrypted, and it would be that network a laptop would need to join in order to be able to communicate with the camera.

So, this is a really cool new option, but it doesn't lead to computers in the stands doing what you theorize.

Greg McKaskle

[tomy]

so...

if you connect a camera to the crio it will do all the image processing
if you connect it to the d-link there is no image processing

when the camera is connect to the crio, if you connect to the router and then go onto a web browser and go to 192.168.90.254 you can see the camera?

and

why doesn't first then up giant screens and connect to the cameras and put them on the projector cause that would be awesome if they did that

[Alan Anderson]

Originally Posted by tomy

so...

if you connect a camera to the crio it will do all the image processing
if you connect it to the d-link there is no image processing

Not necessarily. You can tell the cRIO software to read a camera connected to a network on port 1. If your camera has a compatible address (e.g. 10.x.y.123), it will be accessible to any image processing you want to do.

when the camera is connect to the crio, if you connect to the router and then go onto a web browser and go to 192.168.90.254 you can see the camera?

You can't do that by default, and you won't be able to do it with any reasonable network settings I can think of. First, the cRIO doesn't do any forwarding between its network ports (though it's not hard to tell it to do so). Second, the D-Link router in Access Point mode creates a network that shouldn't route a 192.168 address to the cRIO. In Bridge mode at a competition event, the field network definitely won't route that address.

[tomy]

two more questions:

1)how do you configure the c++ dashboard program to pick up port one?

2)will the lag be less if you connect the camera to the d-link instead of the crio?

[Greg McKaskle]

1. I'm not aware of a C++ dashboard. The dashboard that ships is written in LV and uses the cRIO form of video. I expect a tutorial shortly to show how to modify it to build the other form. The dashboard language is independent of the cRIO language, and you can pretty easily use LV or Java with what is provided. To use C++, be sure to use tools that can produce a Windows executable.

2. The lag will be less, and the cRIO CPU will be available for other things.

Greg McKaskle

[Greg McKaskle]

If you connect the camera to a dlink, there doesn't have to be image processing, but like any computer, the cRIO can connect to the camera and request images.

If you connect to the cRIO, only the cRIO can request images.

It is far easier to connect to the camera with a browser if it is on the dlink. When it is plugged into the cRIO, it is private. There are technically some ways to change routing tables, but now you don't need to do those things.

If FIRST fields start to connect to your camera and stream images, there is some overhead, and it will interfere with your robots access to the images. There isn't a ton of interference, but most likely they won't be doing this. It would be pretty cool though.

Greg McKaskle