The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : Advice on Handling a Camera Vision System Calibration Master


jkittle
5th November 2007, 10:52 AM
Can someone give some advice on handle a calibration master?

I have a camera vision system that inspects for the presence of a brass electrical terminal. What we have had to do is make actual defects for the camera to look for. One of the defects is a terminal that is missing and the camera sees the black plastic in its place. What the camera actually sees is the color difference between shiny brass and the black plastic. How do a control the master? As long as the plastic doesn’t break off it will always be good. What do I put in the calibration system for acceptance, measurement, and standard and so on?

BradM
5th November 2007, 11:03 AM
Can someone give some advice on handle a calibration master?

I have a camera vision system that inspects for the presence of a brass electrical terminal. What we have had to do is make actual defects for the camera to look for. One of the defects is a terminal that is missing and the camera sees the black plastic in its place. What the camera actually sees is the color difference between shiny brass and the black plastic. How do a control the master? As long as the plastic doesn’t break off it will always be good. What do I put in the calibration system for acceptance, measurement, and standard and so on?

It sounds like this may not fall under calibration. Your system does not seem as if it is assessing any error in your parts, but rather a Go/No-Go scenario. I might not call it a calibration, as you are not using NIST traceable standards, and no uncertainty is involved.

I would enter it as a performance check/ verification.

jkittle
5th November 2007, 11:08 AM
The acceptance of the part is either good or bad but when I'm asked how it’s inspected I would say with the vision tester. And the vision tester can be set up incorrectly if the masters are not used. So they are used to set acceptance criteria of inspection equipment.

BradM
5th November 2007, 11:20 AM
How are the masters verified? Or are they just verified initially? Does the vision system have analog values from reading the masters, or is that setup Pass/Fail also?

Can you enter two services for the equipment- A calibration for the vision tester (if applicable) and a verification of the camera system?

As always with these situations, semantics will be important. Do you have identified in your system what you call a calibration? Do you allow for a verification?

jkittle
5th November 2007, 11:31 AM
The tester masters are verified visually (that the plastic post is in tact). When the camera looks for the presence of a brass terminal it sees a black post so the part recognition for the presence of a brass terminal is low (not acceptable). So in this case the master is really being used as verification that the tester is set up properly. But I do need to control them some how sense they are very important to my testing process.

chasf
5th November 2007, 12:37 PM
It sound like you are answering your own question. If your master becomes damaged then your inspection would allow bad parts to get through. To insure this does not happen you could write work instructions for setting up with the master and train operators if need be . Then write a procedure for inspecting the master include inspection intervals. You could use a label on the master with a due date on it. Then keep records of the inspection results. Keep it simple but effective.

yodon
5th November 2007, 04:22 PM
To piggyback on what chasf said, your instructions for setting up the master can include a picture since it sounds like the master can be set up obviously correctly.

We have encountered the same situation but the pass/fail criteria was much more subjective. Controlling the masters has worked for our situation. And I also agree with some earlier posts that you do not want to call this "calibration."

Benjamin28
5th November 2007, 05:19 PM
I would treat this the same way I would a controlled document. I wouldn't consider it a calibration. Give the master a control number, be sure it's referenced in your work instructions, and ensure that only the master is available to the operators for that specific process.