Re: plot profile problems with pixel value

Posted by Jonathan Hilmer on Apr 10, 2008; 9:51pm
URL: http://imagej.273.s1.nabble.com/plot-profile-problems-with-pixel-value-tp3696589p3696592.html

I think the problem you're seeing arises from the spatial definitions
and representations of each pixel.  Rather than go through a lengthy
explanation, try using the point selection tool and Measurements
(which will give the same results as Plot Profile and is faster to
check) when zoomed in.  You'll see the effect causing your results,
which actually arises from a very methodical way of processing the
pixelated data.

The slight difference I see between the point selection tool and line
selections is that the line selections seems to go to the nearest
pixel "corner" for a measurement while the point selection is always
to the upper-left.

Jonathan


On Thu, Apr 10, 2008 at 2:14 PM, Rigo Pantoja <[hidden email]> wrote:

> Hi Harry,
>
>  Wayne recommened your suggestion.  It did not work.
>
>  Basically, when you zoom in and draw a line across a set of linear pixels,
>  you get a variable reading of the pixel value if the line does not cross the
>  pixel in the exact center.  If your pixels values are high you might not
>  notice.  However, if you are analyzing single-molecues - where every count
>  matters - this becomes obvious.
>
>  If you follow the thread you can see this effect with a sample image
>  (fluorescentcells), which after you convert it to a 16bit tiff file.
>
>  Thanks,
>
>  Rigo
>
>  On Thu, Apr 10, 2008 at 10:34 AM, Harry Parker <[hidden email]>
>  wrote:
>
>
>
>  > Have you tried deselecting the menu option, Edit->Options->Profile Plot
>  > Options...->Interpolate Line Profiles?
>  >
>  > --
>  > Harry Parker
>  > Senior Systems Engineer
>  > Digital Imaging Systems, Inc.
>  >
>  > ----- Original Message ----
>  > From: Rigo Pantoja <[hidden email]>
>  > To: [hidden email]
>  > Sent: Wednesday, April 9, 2008 2:35:21 PM
>  > Subject: plot profile problems with pixel value
>  >
>  > All,
>  >
>  > The plot profile (Ctrl+K)  is yielding different values for the same
>  > pixel.
>  > I Noticed this when I generated plot profiles for an individual spot along
>  > the x and y axis.   The  peak  pixel  value was different for the comon
>  > interesection.  Also, if you generate a various plot profiles for with
>  > increasing lengths, the peak pixel value changes.  Can this bug be fixed?
>  >
>  > Thanks,
>  >
>  > Rigo
>  >
>  >
>  >
>  >
>  > __________________________________________________
>  > Do You Yahoo!?
>  > Tired of spam?  Yahoo! Mail has the best spam protection around
>  > http://mail.yahoo.com
>  >
>