So, there are a couple of ways you could do that. One is by using structuring elements as indicated in G &amp; W. Another way would be to take a 3D method like described here in the surface approximation section (shameless plug): <a href="http://confactsdatas.inrialpes.fr/IEEE_RSJ_-_IROS_2010_International_Conference_on_Intelligent_Robots_and_Systems___Conference/data/papers/0389.pdf">http://confactsdatas.inrialpes.fr/IEEE_RSJ_-_IROS_2010_International_Conference_on_Intelligent_Robots_and_Systems___Conference/data/papers/0389.pdf</a><div>
<br></div><div>and adapt it for 2D. </div><div><br></div><div>There is nothing in scikits.image that will do this automatically, but it would be a useful feature so I&#39;ll add it to my TODO list.</div><div><br></div><div>
Chris  <br><div><br></div><div><br><br><div class="gmail_quote">On Wed, Apr 20, 2011 at 12:00 PM, jip <span dir="ltr">&lt;<a href="mailto:jeanpatrick.pommier@gmail.com">jeanpatrick.pommier@gmail.com</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">no, just the convex hull of a a 2D image.<br>I found the definition in &quot;Gonzalez &amp; Woods, p545&quot;, but I am not sure to implement the stuff.<br>
<br>jean-patrick<br></blockquote></div><br></div></div>