defect identification and analysis

[Angelhui]

Dear All,

The defects in maganet surface need to be identified and analyzed. The defficult is that contrast between defect and the other part of the magenet is very low.

Below is a picture, and the shallow long dent defect has a dynamic range of about [65,75].

It is very hard to get the area information of the defect because it is very hard to distinguish it from its outside.

Anyone has some good ideas?

Waiting...

[AnalogKid2DigitalMan · Accepted solution]

If defects will be straight, try IMAQ Find Straight Edge, refer to this thread, be sure to look at the image posted near the end.

http://forums.ni.com/ni/board/message?board.id=200&message.id=17643#M17643

You could also try darkfield lighting (low obliqe angle) to accentuate the defect.

Another alternative- place some of that thin green plastic film (forget the correct trade name) that shows magnet field lines over the magnet and image that. Defect areas should show up very prominetly.

-AK2DM

[AnalogKid2DigitalMan]

Here's that green plastic film I spoke of:

http://scientificsonline.com/product.asp?pn=3082237&cmss=magnetic+viewing

-AK2DM

[hemudu]

AK2DM's idea is great!

NI Applications Engineer

[Robert_Eastlund]

One suggestion would be to utilize polarized lighting to enhance the contrast between the rub mark and the undisturbed surface of the magnet.  Here a polarizer is placed in front of the light source, a ring light, and another polarizer is placed in front of the camera's lens.

If you would like to mail a sample of the part to me, I can test the idea here and email you the results.  My address is as follows:

Graftek Imaging Inc.

Robert Eastlund

8900 Shoal Creek 300 B

Austin, TX 78757

Phone: (800) 441-2118 x101

[Angelhui]

Many thanks to AK2DM, the find straight edge method is also useful here!

Also thanks to Robert.