Steve on Image Processing

December 7th, 2006

Seen on the Web: Bilateral filtering in MATLAB

This looks interesting: Medical Image Processing: Bilateral Filtering in MATLAB. It includes MATLAB code.

By the way, we have been experiencing blog database problems in the last 12 hours or so. We are correcting them as soon as we can. In the meantime, you may see error messages on the page, or you may have difficulty leaving comments. We're sorry for any inconvenience.

11:25 am | Posted in Uncategorized | Permalink |

You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

10 Responses to “Seen on the Web: Bilateral filtering in MATLAB”

1. Steve replied on December 7th, 2006 at 2:40 pm :
   

   I believe the blog database problems have all been resolved now.

2. João replied on December 8th, 2006 at 7:10 pm :
   

   The Bilateral Filtering isn’t implemented correctly. It makes no use of the gaussian closeness function (H_c). The output of the similarity function should be

   V = sum(sum(H_c.*sim.*B)) ./ sum(sum(H_c.*sim));

   The resulting image isn’t much different. I guess they were lucky in their choice of image example.

3. Doug replied on December 8th, 2006 at 7:34 pm :
   

   For those working with color images, please see my implementation of Bilateral Filtering on the Matlab Central File Exchange. The provided source directly implements the approach outlined by Tomasi and Manduchi. In addition, it demonstrates a recent application of the method to automatic image abstraction.

4. Bjorn replied on December 9th, 2006 at 7:50 pm :
   

   …and the idea is apparently patent protected (whatever one might think of software patents):

   S.M. Smith. Method for digitally processing images to determine the position of edges and/or corners therein for guidance of unmanned vehicle. UK Patent 2272285. Proprietor: Secretary of State for Defence, UK. 15 January 1997.

5. Christopher replied on December 12th, 2006 at 6:49 am :
   

   Thanks for the correction. I accidently put up an older version where I forgot the multiplication with the closeness function. It should be okay now.

6. sun replied on January 19th, 2007 at 3:39 am :
   

   hi steve
   can i represent a integer number in MATLAB using fewer than 8 bits?
   ex:
   can i represent 15 in 4 bits?

7. Steve replied on January 19th, 2007 at 6:45 am :
   

   Sun - no.

8. Tom replied on March 21st, 2007 at 7:00 pm :
   

   Hi Steve,

   Bilateral filter implemented by Doug uses loops. Is there anyway to vectorise this code? I tried the filter mentioned in this post but it’s even slower.

   Doug’s code - http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=12191&objectType=file

9. Steve replied on March 22nd, 2007 at 6:29 am :
   

   Tom - I’ll have to take a look. I’ll put it on my list of things to investigate for possible blog postings. I won’t be able to get to it for a while, though.

10. Tom replied on March 22nd, 2007 at 4:59 pm :
    

    Thanks Steve,
    I’ve put it up on Matlab newsgroups too. Maybe some readers will come up with something meanwhile.

    Your blogs are excellent by the way.
    Regards,
    Tom

Leave a Reply

These postings are the author's and don't necessarily represent the opinions of The MathWorks.