Defect in chip image.

Finding defective lines?

Answer 1

[Image Analyst]

What is the defect? That it's less wiggly than the others? Assuming they are aligned with the edges of the image, just sum teh image horizontally and look at the spread in y values (lines).

[Eric Nunes]

Hi, thank you for your reply. Could you explain what do you mean by looking at the spread in y values?

[Image Analyst]

verticalProfile = sum(grayImage, 2);

Then threshold at some value and measure the spread. Let's say a line hits at line 130. Well a straight line might go from 128 - 132, but a wiggly one will go from 122 - 138. It wiggles in a range of 16 rather than 4.

[Eric Nunes]

I tired it , all the lines have a spread of 11 rows, so it did not help much.

[Image Analyst]

You removed your image so now no one can help you anymore.