02-01-2007 09:57 AM
02-01-2007 10:55 AM
02-02-2007 07:38 AM
02-05-2007 07:23 AM
Hi there,
Sorry that's not going to work. I assumed you had the acquisition synchronized with the encoder read. (Therefore all you would need to do it know which position applies to which frame, easy if you have a start trigger)
Unfortunately then the only way of doing this would be to synchronize the acquisition with the encoder position. Are you taking the frames every 1/30th (at a timed interval) of a second, or are you triggering the camera every say 50 pulses.
Decimating the encoder position for use as a triggering pulse for the camera would be the best way to do this type of acquisition. This way the frame rate will purely depend on the speed of the conveyor belt.
How are you taking in the encoder position?
How is the frame rate set up?
Thanks
AdamB
02-05-2007 04:37 PM
10-27-2011 04:32 AM
Hello Rob,
I have almost the same application. It is a webinspection where I need to detects defects with a linescan camera. The downweb (longitudinal) coordinate of the defect must be exchanged with a robot that cuts out the specific part. Since the robot is placed 25 m further away downweb, we synchronize both systems with a counterboard that counts each encoder pulse of 0.25 mm. Each m the robot is synchronised with the vision system by exchanging the (linetrigger) encoder position at the time of the m pulse rising edge to re-reference both systems.
The defect position within a page image is the encodercountervalue of the beginning of that specific page + the row position of the defect in the image translated to encoder pulses. Therefore I need to know the encoder value at the time of capturing that particular page very accuratelly. The framegrabber board does provide a (software) signal that generates a callback. Since we do not use a realtime OS, there is non predictable delay between the actual start of capturing the first line of a page to the execution of the callback in which I can read the actual encoder position. The accuracy of the positioning mainly depends on this delay.
Therefore I am very interested if you already found a solution for your problem, if needed with additional hardware?!?
Looking forward to your reply!
Robbert
10-27-2011 05:34 AM
Hello Robbert,
It is well known problem to get trigger information from the image. Not so far ago I've solved this problem with help of feature called "Chunk Data" which is implemented in Basler's camera. In particular Basler's Pilot GigE camera has the ability to add to each frame special block of data. One portion of this data is "Trigger Counter". This counter can be reset programmatically before start. After that the only thing you have to do is to send trigger signal to the camera, get the image and recieve trigger counter value from this image.
So I suppose that functionality I've just described for Basler cameras might be exists in your camera and it would be great to use it.
I hope I helped you somehow!
Nikita.
11-01-2011 06:04 AM
Hi Nikita,
Thanks for your advise.
Unfortunattely, my camera link camera/framegrabber combination doesnt support this kind of feature.
I am afraid I have to use additional hardware..
Cheers,
Robbert