About build KinectFusion

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

About build KinectFusion

johnnyxia
Hi,

I'm new to PCL. The thing interests me is the demo of KinectFusion. I'd like to build the demo and try it myself.

I searched the blog and mail group, but I didn't find a guide to build it. I tried to build the code in pcl-trunk. It seems that the trunk is always changing. Once I saw a build error, I update the code and the error disappears, but then I saw the other error. I tried it 3 times.

Is there anyone here who builds the KinFu from trunk successfully can share a BKM on it?

Second, currently I'm setting up the build environment on Ubuntu Linux. Is KinFu only supporting Windows?

Thanks in advance.

Regards,
Johnny  

Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

Anatoly Baksheev-2
No special guide for kinfu. You need to build PCL. It's possible to build
only 'pcl_gpu_kinfu' module and its dependencies. You can ignore other
compilation errors.

Kinfu supports all platforms where CUDA is supported. But I never tested
it on Mac.

Cheers,
Anatoly

-----Original Message-----
From: [hidden email]
[mailto:[hidden email]] On Behalf Of johnnyxia
Sent: Sunday, June 03, 2012 7:35 PM
To: [hidden email]
Subject: [PCL-users] About build KinectFusion

Hi,

I'm new to PCL. The thing interests me is the demo of KinectFusion. I'd
like to build the demo and try it myself.

I searched the blog and mail group, but I didn't find a guide to build it.
I tried to build the code in pcl-trunk. It seems that the trunk is always
changing. Once I saw a build error, I update the code and the error
disappears, but then I saw the other error. I tried it 3 times.

Is there anyone here who builds the KinFu from trunk successfully can
share a BKM on it?

Second, currently I'm setting up the build environment on Ubuntu Linux. Is
KinFu only supporting Windows?

Thanks in advance.

Regards,
Johnny



--
View this message in context:
http://www.pcl-users.org/About-build-KinectFusion-tp4018892.html
Sent from the Point Cloud Library (PCL) Users mailing list archive at
Nabble.com.
_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

johnnyxia
Finally, I get it compiled successfully. Here I share my practice here in case it will be helpful to other new comers. BTW, is there a wiki for PCL? Wiki might be a good place for such kind of doc.

My environment is a clean Ubuntu 12.04.

Setup build environment:

$sudo apt-get install subversion cmake cmake-curses-gui build-essential libboost-dev libeigen3-dev libflann-dev  libusb-1.0 libvtk5-dev libqhull-dev
$sudo apt-get install libboost-.*1.46-dev

Next is setup the pcl repository source which is needed for installing openni

$sudo add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
$sudo apt-get update
$sudo apt-get install openni-dev

Next is setup the CUDA, which is needed for building gpu features. Download the CUDA toolkit from:
http://developer.nvidia.com/cuda-downloads

Download the code and build it:

$svn co http://svn.pointclouds.org/pcl/trunk pcl-trunk
$cd pcl-trunk
$mkdir build
$ccmake ..
Make sure visualization feature is enabled, BUILD_GPU is on.
$cmake ..
$make pcl_gpu_kinfu
$make kinfu_app

Q: Since kinfu_app is part of gpu_kinfu subsystem, but I don't know why the kinfu_app is not built when I build gpu_kinfu. Anyway, type the make command again can get the app built successfully

Regards,
Johnny



Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

Radu B Rusu
Administrator
Johnny,

We have a dev wiki at http://dev.pointclouds.org/projects/pcl/wiki. Thanks!

Cheers,
Radu.

On 06/08/2012 08:20 PM, johnnyxia wrote:

> Finally, I get it compiled successfully. Here I share my practice here in
> case it will be helpful to other new comers. BTW, is there a wiki for PCL?
> Wiki might be a good place for such kind of doc.
>
> My environment is a clean Ubuntu 12.04.
>
> *Setup build environment:*
>
> $sudo apt-get install subversion cmake cmake-curses-gui build-essential
> libboost-dev libeigen3-dev libflann-dev  libusb-1.0 libvtk5-dev libqhull-dev
> $sudo apt-get install libboost-.*1.46-dev
>
> Next is setup the pcl repository source which is needed for installing
> openni
>
> $sudo add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
> $sudo apt-get update
> $sudo apt-get install openni-dev
>
> Next is setup the CUDA, which is needed for building gpu features. Download
> the CUDA toolkit from:
> http://developer.nvidia.com/cuda-downloads
>
> *Download the code and build it*:
>
> $svn co http://svn.pointclouds.org/pcl/trunk pcl-trunk
> $cd pcl-trunk
> $mkdir build
> $ccmake ..
> Make sure visualization feature is enabled, BUILD_GPU is on.
> $cmake ..
> $make pcl_gpu_kinfu
> $make kinfu_app
>
> Q: Since kinfu_app is part of gpu_kinfu subsystem, but I don't know why the
> kinfu_app is not built when I build gpu_kinfu. Anyway, type the make command
> again can get the app built successfully
>
> Regards,
> Johnny
>
>
>
>
>
> --
> View this message in context: http://www.pcl-users.org/About-build-KinectFusion-tp4018892p4019155.html
> Sent from the Point Cloud Library (PCL) Users mailing list archive at Nabble.com.
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users
_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

johnnyxia
Radu,

I never searched out that wiki page. It will be better that someone can add a link in pointclouds.org, such the Documentation subpage.

Regards,
Johnny
Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

Radu B Rusu
Administrator
Johnny,

Thanks ! I added this to http://www.pointclouds.org/documentation/.

Cheers,
Radu.

On 06/11/2012 06:41 AM, johnnyxia wrote:

> Radu,
>
> I never searched out that wiki page. It will be better that someone can add
> a link in pointclouds.org, such the Documentation subpage.
>
> Regards,
> Johnny
>
> --
> View this message in context: http://www.pcl-users.org/About-build-KinectFusion-tp4018892p4019224.html
> Sent from the Point Cloud Library (PCL) Users mailing list archive at Nabble.com.
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users
_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: About build KinectFusion

RavaSisters
In reply to this post by johnnyxia
But have you installed the dependencies before doing your practice?


johnnyxia wrote
Finally, I get it compiled successfully. Here I share my practice here in case it will be helpful to other new comers. BTW, is there a wiki for PCL? Wiki might be a good place for such kind of doc.

My environment is a clean Ubuntu 12.04.

Setup build environment:

$sudo apt-get install subversion cmake cmake-curses-gui build-essential libboost-dev libeigen3-dev libflann-dev  libusb-1.0 libvtk5-dev libqhull-dev
$sudo apt-get install libboost-.*1.46-dev

Next is setup the pcl repository source which is needed for installing openni

$sudo add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
$sudo apt-get update
$sudo apt-get install openni-dev

Next is setup the CUDA, which is needed for building gpu features. Download the CUDA toolkit from:
http://developer.nvidia.com/cuda-downloads

Download the code and build it:

$svn co http://svn.pointclouds.org/pcl/trunk pcl-trunk
$cd pcl-trunk
$mkdir build
$ccmake ..
Make sure visualization feature is enabled, BUILD_GPU is on.
$cmake ..
$make pcl_gpu_kinfu
$make kinfu_app

Q: Since kinfu_app is part of gpu_kinfu subsystem, but I don't know why the kinfu_app is not built when I build gpu_kinfu. Anyway, type the make command again can get the app built successfully

Regards,
Johnny