Login  Register

Re: Pixel count via RGB range

Posted by Michael Schmid on Sep 10, 2010; 5:59pm
URL: http://imagej.273.s1.nabble.com/Pixel-count-via-RGB-range-tp3686979p3686982.html

Hi John,

ok, I see. I have uploaded a new version of the Versatile Wand that  
can use the foreground (eyedropper) color instead of the color of a  
given pixel.

-- Attribute table: this question is a different topic.
Anyhow, if you save the image as tiff or zip (i.e., zipped tiff), the  
image properties (including the info) are saved with it. There you  
can also store your own data. Is this what you are looking for?
Otherwise, please start a separate thread in the discussion for it.

Michael
________________________________________________________________

On 10 Sep 2010, at 18:14, John Carpenter wrote:

> Thanks Michael, that worked great. I'd love to automate this  
> process but I
> imagine that would be difficult because the user has to select the  
> pixel?
>
> Is there any way to build an attribute table for the image? I've  
> done this
> with a licensed GIS software but I'd rather use something that is open
> source.
>
>
> On Fri, Sep 10, 2010 at 4:05 AM, Michael Schmid  
> <[hidden email]>wrote:
>
>> Hi John,
>>
>> one option might be the Versatile Wand
>>
>>
>> http://imagejdocu.tudor.lu/doku.php?
>> id=plugin:segmentation:versatile_wand:start
>>
>> (the link should be one line).
>>
>> Use Connectedness 'Non-Contiguous', Gradient Tolerance off (0),  
>> Include
>> Holes off.
>>
>> If you select the tool, click on a pixel with the typical color  
>> and then
>> open the Versatile Wand Options (dobule click the tool icon). You  
>> can use
>> preview and play with the tolerance and Color sliders to get the best
>> result.
>> Then, Analyze>Measure will give you the area selected (make sure that
>> 'Area' is selected in Set Measurements).
>>
>> Michael
>> ________________________________________________________________
>>
>>
>> On 8 Sep 2010, at 23:46, John Carpenter wrote:
>>
>>  Hello,
>>>
>>> I'd like to use this software for digital vegetation profiles.  
>>> The idea is
>>> to place a 0.5m x 0.5m white background (drop cloth, poster  
>>> board,etc.) in
>>> a
>>> field plot and take a picture of it from a certain distance. I'd  
>>> then crop
>>> out just the half meter square area to analyze.
>>>
>>> I suspect that the leaves and twigs would have a fairly small RGB  
>>> range,
>>> as
>>> compared to the white background, that I could specify and then have
>>> ImageJ
>>> count the pixels in that range. All that's left is to divide by that
>>> number
>>> by the total pixels to get a percent cover of vegetation.
>>>
>>> If someone could point me in the right direction on how to do  
>>> this with
>>> ImageJ I would be very grateful.
>>>
>>> Thanks!
>>>
>>