Comments on: Hybrid Segmentation

By: leochow

leochow — Tue, 03 May 2011 14:30:28 +0000

Hi,Shawn
I have just read the paper “Hybrid Geodesic Region-Based Curve Evolutions for Image Segmentation.” I’m wondering whether it’s an error that the operator before the second term is “+” in eq(15) or “-” in eq(16). In eq(14), I think you dropped “-” before the integration.

By: Anjali.S

Anjali.S — Thu, 24 Feb 2011 07:36:06 +0000

Hi,I’ve read your paper Localising region based active contours,”and have some doubts.

In the paper of MS Energy part Eq(17)
It looks as:
[(I-u)^2 / Au] – [(I-v)^2 / Av]

And in your matlab code it shows
F = -((u-v).*((I(idx)-u)./Ain+(I(idx)-v)./Aout))
just like
[(I-u)*(v-u) / Au] – [(I-v)*(u-v)] / Av]

I cannot understand how it came like that.
I look forward your relpy.
Thanks in advance!

By: Anjali.S

Anjali.S — Thu, 24 Feb 2011 07:29:00 +0000

Hi,I’ve read your papers“Active Contours Without Edges,”and I have some doubts.

In the paper of MS Energy part Eq(17)
It looks as:
[(I-u)^2 / Au] – [(I-v)^2 / Av]

And in your matlab code it shows
F = -((u-v).*((I(idx)-u)./Ain+(I(idx)-v)./Aout))
just like
[(I-u)*(v-u) / Au] – [(I-v)*(u-v)] / Av]

I cannot understand it.
I look forward your relpy.
Thanks in advance!

By: Shawn Lankton

Shawn Lankton — Fri, 06 Aug 2010 14:21:50 +0000

@mewada You should look at my paper, “Localizing Region-based Active Contours” where I discuss parameter selection in detail.

By: Mewada Hiren

Mewada Hiren — Fri, 09 Jul 2010 09:44:13 +0000

Sir,
I have gone through the code, I learned well. But i found that code is sensitive to parameter selection like mu, lambda etc. So any method , that select the parameter automatically or update the parameter.

By: Xin

Xin — Wed, 09 Jun 2010 14:35:41 +0000

I appreciate you sharing the codes. It’is really useful.

By: Eran

Eran — Wed, 02 Jun 2010 21:07:03 +0000

Hi Shawn,

I appreciate you sharing the codes.

I am wondering whether your localized region-based active contours segmentation (localized_seg.zip) could perform a multi-object segmentation ( like the elephant example in your paper), please?

By: hindi sms

hindi sms — Sun, 11 Apr 2010 07:29:12 +0000

{i have {seen|read} many so called gurus {sharing|telling} their tips and tricks,but {rarely|very rarely} some are good.I can say very {confidently|prodly|easily} that this post was really {good|awesome|informative}|i would definately {love|like} to see some {good|great} {tips|info} from you to {make|earn} money from adsense.as i {feel|really feel} you can help us. and i really do not believe the other online guru's. They {freeze|cheat} us by {saying|telling} we will make money but instead they make more money.|{good|great} informative {posarticle}.I think i {will need to|should} read some other {post|articles} of your's too.By the way frankly speaking i like the way you appraoch the visitors.Definately {attractive post|innovative post} | This is really interesting to read.hope to see some more like this in future.keep it up the researching and great task.cool|I?m really one of those {people|webmsters} earning next to {nothing|no mopney} or rather $1 a day. {Getting|having} good {traffic|vistors} is really a {big|bit} challenge.|I {luv|like} you lol. You have no idea how helpful {that|this} was, this {site|post} rocks man, ROCKS! WOOO.|I?ve been looking into {information|info} on this {very|particular} topic for quite a while and this really covers it! this is a great blog!}

By: hindi sms

hindi sms — Sat, 10 Apr 2010 23:34:26 +0000

By: Shawn Lankton

Shawn Lankton — Fri, 18 Dec 2009 13:48:11 +0000

@Alex, If you check Sparse Field Active Contours you'll find a fully-3D (and very fast) level sets implementation. You'll have to translate the "energy" function into this framework, although I plan to release the codes to do this very soon.