Notices
The Basement Non-Honda/Acura discussion. Content should be tasteful and "primetime" safe.

Any image processing nerds?

Thread Tools
 
Old Sep 20, 2006 | 05:15 PM
  #1  
Red X's Avatar
Red X
Thread Starter
x pǝɹ
 
Joined: Feb 2003
Posts: 6,246
Likes: 1
From: Upstate
Default Any image processing nerds?

This is a long-shot - but anyone in here familiar with Matlab and/or image processing techniques:

specifically - Filters (Gaussian, Laplacian), Edge detection ?
Reply
Old Sep 20, 2006 | 06:22 PM
  #2  
ReVenGe's Avatar
ReVenGe
Registered User
 
Joined: Feb 2004
Posts: 2,629
Likes: 0
Default

well.... not a nerd.... but i've done some filter things..... what you lookin for?
Reply
Old Sep 20, 2006 | 07:19 PM
  #3  
Red X's Avatar
Red X
Thread Starter
x pǝɹ
 
Joined: Feb 2003
Posts: 6,246
Likes: 1
From: Upstate
Default

I need to implement a derivative of Gaussian function for a 2d image. I'm returning the derivative in the x and y directions.. My question is, do the arrays have to be the same size after I convolute the kernel with the matrice in the x and then in the y?

Convoluting the x and y just means convoluting each individual row with the kernel for x, and column for y?
Reply
Old Sep 20, 2006 | 07:23 PM
  #4  
k3ifers's Avatar
k3ifers
k three ifers
 
Joined: Jun 2002
Posts: 42,568
Likes: 4
From: Buffalo, NY
Default

uhhh.. ??
Reply
Old Sep 20, 2006 | 07:25 PM
  #5  
98CoupeV6's Avatar
98CoupeV6
lots and lots of fail
 
Joined: Dec 1999
Posts: 23,004
Likes: 1
From: Deeeeeeeeeeeeeeetroit
Default

Originally Posted by Red X
I need to implement a derivative of Gaussian function for a 2d image. I'm returning the derivative in the x and y directions.. My question is, do the arrays have to be the same size after I convolute the kernel with the matrice in the x and then in the y?

Convoluting the x and y just means convoluting each individual row with the kernel for x, and column for y?
And I just had an aneurism.
Reply
Old Sep 20, 2006 | 07:27 PM
  #6  
ReVenGe's Avatar
ReVenGe
Registered User
 
Joined: Feb 2004
Posts: 2,629
Likes: 0
Default

i dont think so :hsdunno:

that seems like a realy complicated way to use a blur.... are you just trying to mask it?
Reply
Old Sep 20, 2006 | 07:48 PM
  #7  
Red X's Avatar
Red X
Thread Starter
x pǝɹ
 
Joined: Feb 2003
Posts: 6,246
Likes: 1
From: Upstate
Default

Originally Posted by ReVenGe
i dont think so :hsdunno:

that seems like a realy complicated way to use a blur.... are you just trying to mask it?
No I need to soften the image before I can do edge detection.
Reply
Old Sep 20, 2006 | 07:55 PM
  #8  
James3370's Avatar
James3370
aka Heywoodjablowme
 
Joined: Jul 2003
Posts: 4,562
Likes: 0
From: Nashville, TN
Default

Originally Posted by Red X
I need to implement a derivative of Gaussian function for a 2d image. I'm returning the derivative in the x and y directions.. My question is, do the arrays have to be the same size after I convolute the kernel with the matrice in the x and then in the y?

Convoluting the x and y just means convoluting each individual row with the kernel for x, and column for y?

i think if that had been written in mandarin chinesse, i would have almost the same comprehension level of all that

sefl-wned:

LOL
Reply
Old Sep 20, 2006 | 08:33 PM
  #9  
Draconius's Avatar
Draconius
BAZINGA!
 
Joined: Aug 2004
Posts: 21,613
Likes: 0
From: San Diego, CA
Default

Originally Posted by Red X
I need to implement a derivative of Gaussian function for a 2d image. I'm returning the derivative in the x and y directions.. My question is, do the arrays have to be the same size after I convolute the kernel with the matrice in the x and then in the y?

Convoluting the x and y just means convoluting each individual row with the kernel for x, and column for y?
You are definately in need of a nerd bigger than yourself...
Reply
Old Sep 20, 2006 | 08:42 PM
  #10  
MrFatbooty's Avatar
MrFatbooty
Wannabe yuppie
 
Joined: Dec 2000
Posts: 25,918
Likes: 0
From: Madison, WI
Default

You an EE major or somethin?
Reply



All times are GMT -8. The time now is 01:56 AM.