Hallo together,
i have a EA triggered Vision-source but after triggering positiongenerator need PM to get a item target 500 ms. I ve measured with a stop watch over getitmtgt, start before and stop after getitmtgt.
Is this normal or can i make this faster.
Thanks, Martin
Hi,
Maybe your vision-model needs a long time to process?
In the training window where you configure your vision model you can see how long the processing time of your vision model is. If you are searching for more than one model in every image you need to add the processing times for each model together to get the total required time.
//Fredrik
Hi,
the Vision-Model need 20 ms, Calibration-Corr. 10 ms, Camera-Shutter 4 ms.
I have an ind. Workarea with the Camera. From Trigger to getitmtgt-answer PM need
more than 500ms. Is this correct? If i use a CNV is this not a Problem, because i have the
time between trigger and drive Product from Camera to Workarea.
Has anybody a solution.
Martin
Several years ago, we delivered a Pickmaster application with 4 Basler cameras and 4 robots. We were feeding product from several sources to a fixed location area under a camera. Whenever product arrived, we would trigger the camera. Our model processing times were higher, and our delays were much less. By memory, I would estimate that the targets were available at the robot shortly after the theoretical processing time was complete - maybe a 100ms delay or even less.
The only issue we experienced was that the system was not reliable if the camera triggers were too frequent. We had to trigger the Basler Scout’s at less than half of their rated frame rate on any two subsequent triggers.I don’t recall taking any special steps to achieve the quick throughput of data. I’m sorry if this doesn’t help, but I know Pickmaster was capable of better performance than you are observing and it shouldn’t be a fundamental limit of the software.
Martin, since all the signals are passing through the robot controller, (position generator, vision trigger, camera strobe), I recommend you to check some different times:
Difference between position generator and vision trigger, difference between vision trigger and camera strobe, difference between camera strobe and result available in rapid.
/Mats