Image streaming delays from vision sensor images acquisition
Posted: 12 Jul 2013, 13:43
hi, I'm acquiring a video streams from a group of vision sensors (cameras) in V-Rep.
I use to read the frames through some ROS topics and, afterwards, elaborate the images using the OpenCV library.
After some image processing i want to perform some position control on the sensors.
when I was using only 2/3 sensors their movements were correct and fluent.
Now I'm trying to use more sensors (actually 8) and the images are acquired in a discontinuous way: there are some delays in the communication that happen randomically , last randomically and not always for the same sensors.
In fact if i try to display the video streams through a ROS node i see them sometime stopping at a certain frame and starting again after some seconds (visualizing different images because the camera moved in the meantime).
Is there some solution to that.?..i can accept that the acquisition could be not performed in real time but I need it continuous!
Thanks!
I use to read the frames through some ROS topics and, afterwards, elaborate the images using the OpenCV library.
After some image processing i want to perform some position control on the sensors.
when I was using only 2/3 sensors their movements were correct and fluent.
Now I'm trying to use more sensors (actually 8) and the images are acquired in a discontinuous way: there are some delays in the communication that happen randomically , last randomically and not always for the same sensors.
In fact if i try to display the video streams through a ROS node i see them sometime stopping at a certain frame and starting again after some seconds (visualizing different images because the camera moved in the meantime).
Is there some solution to that.?..i can accept that the acquisition could be not performed in real time but I need it continuous!
Thanks!