I guess this is related to issue #6 The scenario is that I implement a service that sends some information to the camera node and runs the image grep loop till I have one image. I trigger the camera externally and call this service. In this case it needs to check for an image at leased 3 times, with an timeout of 2000ms to get an image. If I minimize the timeout to 40ms I still need to run "get image" around 3 times and I get the image earlier cause of the smaller timeout. Also, I do not trigger again during the image readout.
This tells me that the driver is not really checking for images "during" the timeout.
issue #6 has some actionable suggestions on how to reduce USB bandwidth... if those don't solve the issue, then it might not be related
"implement a service" -> ROS service server on your own node? or generic RPI service?
"image grep loop"?
"not really checking for images during the timeout": not sure what you mean, but as long as there's a subscriber connected, the ueye_cam nodelet will actively attempt to grab frames
Hello,
I guess this is related to issue #6 The scenario is that I implement a service that sends some information to the camera node and runs the image grep loop till I have one image. I trigger the camera externally and call this service. In this case it needs to check for an image at leased 3 times, with an timeout of 2000ms to get an image. If I minimize the timeout to 40ms I still need to run "get image" around 3 times and I get the image earlier cause of the smaller timeout. Also, I do not trigger again during the image readout.
This tells me that the driver is not really checking for images "during" the timeout.