On 23 November 2011 13:43, Adrián del Pino <delpinonavarr...@gmail.com> wrote:
>
>
> On Tue, Nov 22, 2011 at 10:42 PM, Carnë Draug <carandraug+...@gmail.com>
> wrote:
>>
>> On 22 November 2011 20:53, Adrián del Pino <delpinonavarr...@gmail.com>
>> wrote:
>> > On Mon, Nov 21, 2011 at 9:59 PM, Carnë Draug <carandraug+...@gmail.com>
>> > wrote:
>> >>
>> >> On 26 October 2011 16:19, Adrián del Pino <delpinonavarr...@gmail.com>
>> >> wrote:
>> >> > Dear Octave:
>> >> > Sometimes the function bweuler returns incorrect values. By
>> >> > example,
>> >> > it
>> >> > returns -0.5 instead of 1 with this image ("prueba3.tif"). Moreover,
>> >> > this
>> >> > function should return integer values.
>> >> >
>> >> > Best regards.
>> >> > Adrián del Pino
>> >>
>> >> Hi Adrián
>> >>
>> >> just to let you know that this was not forgotten. I was hoping that
>> >> someone would show up and manage to fix it. As that doesn't seem to be
>> >> happening, I added your report to the bug tracker here
>> >>
>> >>
>> >> https://sourceforge.net/tracker/?func=detail&aid=3440926&group_id=2888&atid=102888
>> >>
>> >> It would be great if you were able to fix the problem and submit a
>> >> patch for it. The function is written in the GNU octave language so it
>> >> shouldn't be too hard provided you know the subject. The link for the
>> >> latest revision of the file is
>> >>
>> >>
>> >> http://octave.svn.sourceforge.net/viewvc/octave/trunk/octave-forge/main/image/inst/bweuler.m
>> >>
>> >> I have also CC'ed the dev that wrote this function.
>> >>
>> >> Carnë
>> >
>> > Hi Carnë,
>> > I have fixed two problems in this function. There was no validation
>> > of
>> > the input image and there was a problem computing values when a figure
>> > touches the top and left borders.
>> >
>> > I want to submit the new version but the procedure seems a little
>> > complex.
>>
>> You are right. I should spend an hour sometime this week and write
>> more up to date and simpler instructions. It's actually quite simple.
>>
>> > Also, I don't know if I should write the log at the end of the file.
>>
>> No, you should not. You did right.
>>
>> > I have attached the file to receive your comments.
>>
>> I made a few more changes but mostly aesthetical. I defined the
>> defaults of n at the top with "function eul = bweuler (BW, n = 8)" and
>> used print_usage for wrong inputs. I made sure that n was a numeric
>> scalar before checking its value otheriwse some weird error would
>> appear if the user tried to use a string.
>>
>> I also changed a bit the help text, and added you to copyright notice.
>> Is this the e-mail you want on the copyright notice? See the attached
>> file.
>>
>> Yes, it is.
>>
>>>
>>> Also, the help text mentions that it uses the Bit Quads as described
>>> in "Digital Image Processing". I don't know if this is still true
>>> after your changes.
>>>
>> Yes. It continues true. I only put some zeros in the image borders
>> to capture the values that applylut returns for the border pixels
>> (applylut put these values in the top-right corner of each two-pixel
>
> square).
>
> It doesn't look as described in the book because someone changed
> the original formulas to save computation time. There was an
> explanation
> in the previous version that I return to the file. However, I
> changed this
> method because the real computations are slower than the integer
> ones.
>
>>> If no one objects against the attached file in 3 days, I'll commit it.
>>>
>>> Carnë
>
> Best regards,
> Adrián
>
Cool. Thank you for fixing this. I have just comitted it and it will
be on the next release of the image package.
Carnë
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure
contains a definitive record of customers, application performance,
security threats, fraudulent activity, and more. Splunk takes this
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Octave-dev mailing list
Octave-dev@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/octave-dev
