Separate defects Short,Spur and Spurious Copper based on tracking the boundary of the defects on all directions.

Hello,

I am working on classification of defects on PCB using matlab. I am considering 1 reference image and test image..By subtracting the two images,output consist of 3 defects. SHORT SPUR SPURIOUS COPPER I wanted to separate these defects Short,Spur and spurious copper based on tracking the boundary of the defects on all four sides(North South East West)on the test image. since .....short is always connected between two conductors. .....spur is overlaid on one side of the conductor .....spurious copper is not having any connectivity with other conductor. so based on vicinity tracks i want to classify the defects. ie.. short- 2 vicinity tracks spur - 1 vicinity track spurious copper- 0 vicinity track

Requesting you to please help me in writing matlab program to separate this 3 defects. Attaching herewith matlab code,reference image and test image.

Thanks and Regards,

Sagar Kundri

[Jan]

What kind of help do you want? Did you try to start already? If so, what have you done so far?

[Sagar Kundri]

Hi Jan,

I am writing matlab code to implement multiple defect classifications in hardware pcb. I have already completed 10 defects which I need to.

Now I am stuck at the next step where we have defects such as Spur, Short and Spurious Copper. I need to get this defects seperated into seperate images. To get the output, I have one reference image and another image from which we need to get the defects images. I have a logic to implement, but not sure how to proceed with same. The logic would be to compare the 2 images based on the North direction of both the images.

I am not sure how to attach the images here for reference so that you could also have a look at the same. But this is the description in short about what I want to achieve. Let me know if there are any questions.

Regards,

Sagar Kundri

[Sagar Kundri]

Hi,

Is there any update regarding my query or anyone can help me here. For your reference I have mentioned the pcb images weblinks below for your reference. I request for help as this is urgent.

1. Reference Image

http://tinypic.com/view.php?pic=2yy2fzn&s=5

2. Image with Defects.

http://tinypic.com/view.php?pic=90x3i8&s=5

If anyone has any questions, pease feel free to ask.

Regards,

Sagar Kundri

[Walter Roberson]

http://www.mathworks.com/matlabcentral/answers/29922-why-your-question-is-not-urgent-or-an-emergency