# camera client node for voxl-dfs-server or voxl-qvio-server

Source: https://forum.modalai.com/topic/358/camera-client-node-for-voxl-dfs-server-or-voxl-qvio-server
Category: General Questions (https://forum.modalai.com/category/2/general-questions)
Posted: 2021-07-27 09:13:11 UTC by shlee853
Replies: 2 · Views: 708

## shlee853 · 2021-07-27 09:13:11 UTC

Hi dev team,

I'm investigating source code for qvio, dfs-server.

As you can see the main code, it broadcast their image processing result using modal pipe server.

But, in the client code, it only get data result using modal pipe client, even it generates image stream.

1) How can I get streaming result for point cloud or qvio from server?

My meaning is not steal image, image stream like rtsp client.

2) If there are a few types of camera server such as camera node from voxl-camera-server result, point-cloud result from voxl-dfs-server or qvio result from voxl-qvio-server, how can client recognize which camera server matched with?

Thanks,

## Reply by Guest · 2021-07-27 16:22:41 UTC

Hi Shlee,

Much like camera server, these two services output data to /run/mpa . If you're running our latest voxl suite release (0.4.6), you can run 
```
voxl-streamer -c qvio-overlay
```
 if you want an rtsp stream of the HUD coming out of qvio server or 
```
voxl-streamer -c dfs-disparity
```
If you want the disparity image from DFS server.

If you're not on our latest release (though we strongly encourage that you do install the platform 033-046 [here](https://developer.modalai.com/asset)) then you can manually add voxl-streamer configurations to the config file as seen [here](https://gitlab.com/voxl-public/utilities/voxl-streamer/-/blob/master/config/voxl-streamer.conf#L53)

If you want raw qvio data or the DFS pointcloud then you'll need to do that through [voxl-mpa-to-ros](https://gitlab.com/voxl-public/ros/voxl_mpa_to_ros) which will publish the two aforementioned images, as well as a DFS pointcloud and raw qvio data.

## Reply by shlee853 · 2021-07-27 19:40:46 UTC

Great,

Thank you for your reply.