Dawn Robotics Forum Support and community forums for Dawn Robotics Ltd 2014-09-29T14:51:57+01:00 http://forum.dawnrobotics.co.uk/feed.php?f=11&t=1282 2014-09-29T14:51:57+01:00 2014-09-29T14:51:57+01:00 http://forum.dawnrobotics.co.uk/viewtopic.php?t=1282&p=1466#p1466 <![CDATA[Re: Raspberry Pi Camera Robot Kit Question]]>
Welcome to the forums. :) I'll answer this question, and then send a reply to your other question.

The example programs that are provided with the py_websockets_bot library will run both on an external PC, and on a Pi. When running the image processing and motion vector samples on the Pi, you will need to run them from the desktop to begin with, as they make use of some of the OpenCV windowing functionality to display their output (cv2.imshow etc). If remove calls to imshow etc from your programs then they will also run without an X server.

Now, having said that. You can run all of your image and sensor processing code on the Pi, but the Pi doesn't have a very powerful CPU, so it can be slow, especially when performing image processing. To help make things a bit quicker, the py_websockets_bot library provides the option to stream 'small' (160x120) images that have been scaled down on the GPU.

Hope that helps.

Regards

Alan

Statistics: Posted by Alan — Mon Sep 29, 2014 2:51 pm


]]>
2014-09-27T14:58:52+01:00 2014-09-27T14:58:52+01:00 http://forum.dawnrobotics.co.uk/viewtopic.php?t=1282&p=1460#p1460 <![CDATA[Raspberry Pi Camera Robot Kit Question]]>
I posted an unrelated question in the Rover 5 forum earlier but I have a question about this kit.

If I was to follow the OpenCV image processing tutorial and adding a sensor tutorial to add that functionality would this bot be completely independent in the sense that it would be able to process the images for itself and not need a PC or any user input to take pictures, process and move.

Thanks

Statistics: Posted by Medelwr — Sat Sep 27, 2014 2:58 pm


]]>