A Few More Steps for Machine Vision Camera Evaluation and Selection

Posted by Gretchen Alper on Thu, Jan 17, 2013

Once you have completed the steps to prepare for your machine vision camera evaluation, you are ready to receive the cameras and begin testing.

 Step 6:  Test the Cameras

There are many different ways to test the cameras and the right way for you depends on the most important parameters for your measurement.

One place to start is an image in dark.  This can help you determine the noise (fixed pattern noise), for instance.  The EMVA 1288 can serve as a guideline for consistent testing procedures for certain specifications if they are pertinent to your system. 

The Photon Transfer Curve can allow for a thorough quantitative analysis and derivation of several parameters.  In order to generate it, special equipment is required (an integrating sphere).  

If you are short on time or test equipment, there are independent testing houses that can be a good resource.  OR, many camera manufacturers have Photon Transfer Curves and can provide you with the data for analysis.  They can provide other detailed data as well, such as sensitivity at a specific wavelength.  This can help you minimize your time required, but still perform a thorough quantitative analysis.


Step 7:  Share Results

We cannot emphasize enough how beneficial it can be to share your results with the corresponding camera manufacturer.  Whether the results are good or bad, this can allow an open discussion for troubleshooting and explanations before any final conclusions are made.  Also there may be options for the camera provider to make (small) changes to the camera or production process (such as match acceptance testing to incoming inspection testing) for you that would solve other problems once they understand your challenges better.


Step 8:  Narrow the List and Test Multiple Copies for Final Selection

When you have narrowed down to say the top 2 options, ask the camera vendors to send multiple copies of the camera or perform measurements on multiple samples for you.  With these, you can evaluate the camera consistency and check the repeatability of your results.  This helps you create a final solution with the vendor or create a range of adjustments to compensate elsewhere in your imaging system depending on your choice.


At the end of all this, you should have a good idea of what to expect in your system and hopefully close partnership with the camera vendor.  Best of luck in your search!


Topics: Vision System Optimization