percent cover

[binoink]

Hello,

I am trying to use image J to calculate percent cover and or percent green.
I have overhead photos of vegetation transects at the beach and need to
calculate what amount of a 1x1m area is covered by vegetation.  The sand is
white-ish and all plants are green.  Any suggestions or a protocol on how to
do this?

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[gankaku]

Hi Bianca,

To be able to help you better it would be good if you could attatch a
sample image here. This makes it easier to check which method might be
applicable in your case

What you definitely need to do is to extract the "plant" information from
the rest. There are different possibilities.

1.) you can manually threshold your image (which I guess is an RGB true
color image): > Image >Adjust >Threshold. This will directly guve you the
chance to extract the leaf information by adjusting the sliders.
Potentially the best output you will get by using the HSB color space for
extraction and limit using th esliders to specific colors and saturations.
The threshold determined in one image by this approach might not be equally
applicable to all images in the worst case.

2.) If plant and background are sufficiently distinguishable it might be
easier to transfer the image to grayscale (>Image >Type >8-bit).
Potentially, you might want to check first if you want to do this weighted
to get a better separation and less loss of information (>Edit >Options
>Conversions... --> activate the weighted conversion checkbox first). The
latter includes color information while transferring the image to grayscale.
Thereafter, it might help to find a Auto Threshold (
http://fiji.sc/Auto_Threshold) to extract the desired information (>Image >
Adjust >Auto Threshold). Those might be rather applicable to different
images than the manual ones.
If you have difficulties in deciding which Auto Threshold is suitable you
might want to install the BioVoxxel Toolbox (
http://fiji.sc/BioVoxxel_Toolbox) and use the Threshold Check (
http://fiji.sc/BioVoxxel_Toolbox#Threshold_Check) which makes this decision
a bit easier.

3.) In a very difficult case you might want to think about using a
trainable segmentation tool like a) SIOX (>Plugins >Segmentation >SIOX:
Simple Interactive Object Extraction or b) even the Trainable Weka
Segmentation (>Plugins >Segmentation >Trainable Weka Segmentation).

After the application of one of those methods you will get a binary image
with your information of interest in white on a black background or vice
versa (this can be determined under >Process >Binary >Options... --> the
black background checkbox). Just make sure that you are consistent with the
latter setting to not finally ending up in analyzing your background
instead of your plants (see:
http://rsbweb.nih.gov/ij/docs/guide/146-29.html#sub:BinaryOptions...)

Thre rest is traight forward. You can run a measurement (>Analyze
>Measure). This will give you the results table which holds the information
"%Area". This area fraction is what you are looking for. If this is not
shown you might first need to determine the measurements under >Analyze
>Set Measurements and activate at least the "Area fraction" checkbox.

Once you found a working method you might also want to record this to not
always do it manually (>Plugins >Macros >Record...)

Hope this helps you in getting forward.

Kind regards,
Jan


2014-07-03 21:17 GMT+02:00 Bianca Reo <[hidden email]>:

> Hello,
>
> I am trying to use image J to calculate percent cover and or percent green.
> I have overhead photos of vegetation transects at the beach and need to
> calculate what amount of a 1x1m area is covered by vegetation.  The sand is
> white-ish and all plants are green.  Any suggestions or a protocol on how
> to
> do this?
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--

BIOVOXXEL
CEO: Dr. rer. nat. Jan Brocher
phone:  [phone]
mobile: [phone]
e-mail: [hidden email]
info: [hidden email]
inquiries: [hidden email]
web: www.biovoxxel.de

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html