<div dir="auto"><div>Laurent, list,<br><br><div class="gmail_quote gmail_quote_container"><div dir="ltr" class="gmail_attr">On Tue, Mar 4, 2025, 06:06 celati Laurent via QGIS-User <<a href="mailto:qgis-user@lists.osgeo.org">qgis-user@lists.osgeo.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">
<p>Hello.<br>
My goal is to generate a relevant segmentation that would delimit the
main physiopnomic units (homogenous areas). Up to now, I have done
several tests with qgis and orfeotoolbox 9 with the OTB Generic Region
merging module : <a href="https://www.orfeo-toolbox.org/CookBook-7.0/Applications/app_GenericRegionMerging.html" rel="noopener nofollow ugc noreferrer" target="_blank">GenericRegionMerging — Orfeo ToolBox 7.0.0 documentation</a> with those folowing settings</p>
<ul><li>input image : spot 6/7 satelite image (band blue, green, red, near infrared)<br>
or Aerial french mapping agency images (50 cms bd ortho with green, red and near infrared).<br>
*Homogenity criterion to use : Baatz & Schape</li><li>Threshold for the criterion : 500</li><li>Number of iterations`: 0</li><li>Weight for the spectral homogeneity : 0.1</li><li>Weight for the spatial homogeneity : 0.1</li></ul>
<p>The result is rather interesting for certain areas, in particular
crops areas, without too much vegetation. on the other hand, the method
does not seems optimal, in particular for forest areas.<br>
I’m taking the liberty to paste some screenshots :<br>
the first SS showing interesting result on crop areas :</p>
<br><div><img src="cid:ii_m7uka7lh0" alt="4d3f841e567bd5ac5ce37035342c4e71e78d16c6.jpeg" width="257" height="194" style="margin-right:0px"></div><div>
A second SS showing the non relevant result on forest areas : <br></div><div><img src="cid:ii_m7ukaojr1" alt="9b64943de826eddd230762b3d2a5eb358cc423ec.jpeg" width="260" height="143" style="margin-right:0px"></div><div>
Perhaps for these forest areas, the color/spectral criterion is not
relevant. It would be better to generate / integrate texture
information? As far as i understood, this GRM OTB module is based only
on spectral/colour criterion but not take into consideration texture
information.<br>
Are there other orfeotoolbox segmentation module that takes into account
not only the color criterion (spectral) but also texture information?<br>
If this is not the case, should i create some kind of new raster (add a
band to my existing color infrared raster) dedicated to texture and/or
ndvi information?<br>
Thanks so much for your guidances.
<br></div></div></blockquote></div></div><div dir="auto"><br></div><div dir="auto">Laurent, you don't give us an idea of what forest properties you're hoping to differentiate.</div><div dir="auto"><br></div><div dir="auto">Forests tend to have a vertical vegetation structure, especially natural forests; often with dominant species overtop of second and even third tree species, with other vegetation layers beneath, including shrubs, grasses and bryophytes going on down to the litter and soil.</div><div dir="auto"><br></div><div dir="auto">In my experience, capturing this kind of complex structure using only reflectance is difficult. LIDAR can get at least some of this vertical structure, with some ability to differentiate tree species based on previously known local height and crown shape characteristics.</div><div dir="auto"><br></div><div dir="auto">But maybe that's more than you're after!</div><div dir="auto"><br></div><div dir="auto">This paper might be of interest to you;</div><div dir="auto"><br></div><div dir="auto"><a href="https://www.orfeo-toolbox.org/orfeo-toolbox-and-geobia-classification-of-uav-imagery-in-forest-ecosystems/">https://www.orfeo-toolbox.org/orfeo-toolbox-and-geobia-classification-of-uav-imagery-in-forest-ecosystems/</a></div><div dir="auto"><br></div><div dir="auto">Good luck.</div></div>