Conversion from dense to not dense cloud.

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

Conversion from dense to not dense cloud.

luca
Is there any simple method for converting a dense cloud to a standard,
not dense, pcl cloud, without directly copying all data? any dedicated
method?

Thanks

Luca

_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: Conversion from dense to not dense cloud.

aichim
Administrator
Hi,

What do you define by dense and not dense?

Cheers,
Alex
---
http://www.openperception.org



On Mar 19, 2013, at 4:27 PM, Luca Penasa <[hidden email]> wrote:

> Is there any simple method for converting a dense cloud to a standard, not dense, pcl cloud, without directly copying all data? any dedicated method?
>
> Thanks
>
> Luca
>
> _______________________________________________
> [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: Conversion from dense to not dense cloud.

koen buys-2
Luca,

You might want to read up on the definition we handle for "dense":

http://www.pointclouds.org/documentation/tutorials/basic_structures.php#basic-structures

A dense pointcloud is a subgroup of the "not dense" pointclouds.
So try to state your problem differently, so we can try to help you out.

Best regards,

Koen



On 20 March 2013 00:47, Alexandru Ichim <[hidden email]> wrote:
Hi,

What do you define by dense and not dense?

Cheers,
Alex
---
http://www.openperception.org



On Mar 19, 2013, at 4:27 PM, Luca Penasa <[hidden email]> wrote:

> Is there any simple method for converting a dense cloud to a standard, not dense, pcl cloud, without directly copying all data? any dedicated method?
>
> Thanks
>
> Luca
>
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users

_______________________________________________
[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: Conversion from dense to not dense cloud.

luca
Sure! You are right, I stated the problem in the wrong way. in fact I completely misunderstood the meaning of dense.

So... What i meant is: how can I convert from an organized to a non-organized point cloud?
from a cloud with "height=m, width=n" to a cloud with "height = 1, width=mxn". Is there any dedicated filter or method?

I wrongly thought that the dense bool was referred to this difference.
Thank you for pointing me to the documentation page :-)

Many thanks

Luca
 

On 03/20/2013 09:00 AM, Koen Buys wrote:
Luca,

You might want to read up on the definition we handle for "dense":

http://www.pointclouds.org/documentation/tutorials/basic_structures.php#basic-structures

A dense pointcloud is a subgroup of the "not dense" pointclouds.
So try to state your problem differently, so we can try to help you out.

Best regards,

Koen



On 20 March 2013 00:47, Alexandru Ichim <[hidden email]> wrote:
Hi,

What do you define by dense and not dense?

Cheers,
Alex
---
http://www.openperception.org



On Mar 19, 2013, at 4:27 PM, Luca Penasa <[hidden email]> wrote:

> Is there any simple method for converting a dense cloud to a standard, not dense, pcl cloud, without directly copying all data? any dedicated method?
>
> Thanks
>
> Luca
>
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users

_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users



_______________________________________________
[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: Conversion from dense to not dense cloud.

koen buys-2
The date is stored in an std::vector<PointT>, so the only indication for the organized vs non-organized is the width and height field.
If you want the pointcloud to be unorganized, just change those fields to "height = 1, width=mxn".

However why would you want to do this?

Koen



On 20 March 2013 11:09, Luca Penasa <[hidden email]> wrote:
Sure! You are right, I stated the problem in the wrong way. in fact I completely misunderstood the meaning of dense.

So... What i meant is: how can I convert from an organized to a non-organized point cloud?
from a cloud with "height=m, width=n" to a cloud with "height = 1, width=mxn". Is there any dedicated filter or method?

I wrongly thought that the dense bool was referred to this difference.
Thank you for pointing me to the documentation page :-)

Many thanks

Luca

 

On 03/20/2013 09:00 AM, Koen Buys wrote:
Luca,

You might want to read up on the definition we handle for "dense":

http://www.pointclouds.org/documentation/tutorials/basic_structures.php#basic-structures

A dense pointcloud is a subgroup of the "not dense" pointclouds.
So try to state your problem differently, so we can try to help you out.

Best regards,

Koen



On 20 March 2013 00:47, Alexandru Ichim <[hidden email]> wrote:
Hi,

What do you define by dense and not dense?

Cheers,
Alex
---
http://www.openperception.org



On Mar 19, 2013, at 4:27 PM, Luca Penasa <[hidden email]> wrote:

> Is there any simple method for converting a dense cloud to a standard, not dense, pcl cloud, without directly copying all data? any dedicated method?
>
> Thanks
>
> Luca
>
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users

_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users



_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users


_______________________________________________
[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: Conversion from dense to not dense cloud.

luca

On 03/20/2013 11:30 AM, Koen Buys wrote:
The date is stored in an std::vector<PointT>, so the only indication for the organized vs non-organized is the width and height field.
If you want the pointcloud to be unorganized, just change those fields to "height = 1, width=mxn".

However why would you want to do this?
Some time ago i wrote the qPCL plugin for cloudcompare:
(https://github.com/cloudcompare/trunk/tree/master/qCC/plugins/qPCL)

in which I created a class for converting pcl' pointclouds to cloudcompare internal format :
(https://github.com/cloudcompare/trunk/blob/master/qCC/plugins/qPCL/PclUtils/utils/sm2cc.cpp).

It seems to work fine but when I try to open a structured pcd if fails.
I should check what I wrote but I was wondering if it could depend by some internal difference between structured/non structured clouds that I did not consider when I wrote that class.

Thanks Koen!

Koen



On 20 March 2013 11:09, Luca Penasa <[hidden email]> wrote:
Sure! You are right, I stated the problem in the wrong way. in fact I completely misunderstood the meaning of dense.

So... What i meant is: how can I convert from an organized to a non-organized point cloud?
from a cloud with "height=m, width=n" to a cloud with "height = 1, width=mxn". Is there any dedicated filter or method?

I wrongly thought that the dense bool was referred to this difference.
Thank you for pointing me to the documentation page :-)

Many thanks

Luca

 

On 03/20/2013 09:00 AM, Koen Buys wrote:
Luca,

You might want to read up on the definition we handle for "dense":

http://www.pointclouds.org/documentation/tutorials/basic_structures.php#basic-structures

A dense pointcloud is a subgroup of the "not dense" pointclouds.
So try to state your problem differently, so we can try to help you out.

Best regards,

Koen



On 20 March 2013 00:47, Alexandru Ichim <[hidden email]> wrote:
Hi,

What do you define by dense and not dense?

Cheers,
Alex
---
http://www.openperception.org



On Mar 19, 2013, at 4:27 PM, Luca Penasa <[hidden email]> wrote:

> Is there any simple method for converting a dense cloud to a standard, not dense, pcl cloud, without directly copying all data? any dedicated method?
>
> Thanks
>
> Luca
>
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users

_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users



_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users


_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users




_______________________________________________
[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: Conversion from dense to not dense cloud.

Andreas Mützel
On Wednesday 20 March 2013, 12:37:37 Luca Penasa wrote:

> On 03/20/2013 11:30 AM, Koen Buys wrote:
> > The date is stored in an std::vector<PointT>, so the only indication
> > for the organized vs non-organized is the width and height field.
> > If you want the pointcloud to be unorganized, just change those fields
> > to "height = 1, width=mxn".
> >
> > However why would you want to do this?
>
> Some time ago i wrote the qPCL plugin for cloudcompare:
> (https://github.com/cloudcompare/trunk/tree/master/qCC/plugins/qPCL)
>
> in which I created a class for converting pcl' pointclouds to
> cloudcompare internal format :
> (https://github.com/cloudcompare/trunk/blob/master/qCC/plugins/qPCL/PclUtils
> /utils/sm2cc.cpp).
>
>
> It seems to work fine but when I try to open a structured pcd if fails.
> I should check what I wrote but I was wondering if it could depend by
> some internal difference between structured/non structured clouds that I
> did not consider when I wrote that class.

Hi,

often organized clouds contain NaN values, e.g. when using Kinect data which
has "holes" where no depth value is measured.
Maybe those cause problems in your case?

Cheers,
Andreas


_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: Conversion from dense to not dense cloud.

luca
It could be.
In my case I perform a fromROSMsg that seems to return a cloud with
invalid coordinates, but I should investigate better this problem.

Is there a filter for removing NaNs from PointCloud2 format?

Thanks

On 03/20/2013 11:48 AM, Andreas Mützel wrote:

> On Wednesday 20 March 2013, 12:37:37 Luca Penasa wrote:
>> On 03/20/2013 11:30 AM, Koen Buys wrote:
>>> The date is stored in an std::vector<PointT>, so the only indication
>>> for the organized vs non-organized is the width and height field.
>>> If you want the pointcloud to be unorganized, just change those fields
>>> to "height = 1, width=mxn".
>>>
>>> However why would you want to do this?
>> Some time ago i wrote the qPCL plugin for cloudcompare:
>> (https://github.com/cloudcompare/trunk/tree/master/qCC/plugins/qPCL)
>>
>> in which I created a class for converting pcl' pointclouds to
>> cloudcompare internal format :
>> (https://github.com/cloudcompare/trunk/blob/master/qCC/plugins/qPCL/PclUtils
>> /utils/sm2cc.cpp).
>>
>>
>> It seems to work fine but when I try to open a structured pcd if fails.
>> I should check what I wrote but I was wondering if it could depend by
>> some internal difference between structured/non structured clouds that I
>> did not consider when I wrote that class.
> Hi,
>
> often organized clouds contain NaN values, e.g. when using Kinect data which
> has "holes" where no depth value is measured.
> Maybe those cause problems in your case?
>
> Cheers,
> Andreas
>
>
> _______________________________________________
> [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: Conversion from dense to not dense cloud.

Adam Stambler
You can use a passthrough filter to remove NaN points in one step.



On Wed, Mar 20, 2013 at 8:16 AM, Luca Penasa <[hidden email]> wrote:
It could be.
In my case I perform a fromROSMsg that seems to return a cloud with invalid coordinates, but I should investigate better this problem.

Is there a filter for removing NaNs from PointCloud2 format?

Thanks


On 03/20/2013 11:48 AM, Andreas Mützel wrote:
On Wednesday 20 March 2013, 12:37:37 Luca Penasa wrote:
On 03/20/2013 11:30 AM, Koen Buys wrote:
The date is stored in an std::vector<PointT>, so the only indication
for the organized vs non-organized is the width and height field.
If you want the pointcloud to be unorganized, just change those fields
to "height = 1, width=mxn".

However why would you want to do this?
Some time ago i wrote the qPCL plugin for cloudcompare:
(https://github.com/cloudcompare/trunk/tree/master/qCC/plugins/qPCL)

in which I created a class for converting pcl' pointclouds to
cloudcompare internal format :
(https://github.com/cloudcompare/trunk/blob/master/qCC/plugins/qPCL/PclUtils
/utils/sm2cc.cpp).


It seems to work fine but when I try to open a structured pcd if fails.
I should check what I wrote but I was wondering if it could depend by
some internal difference between structured/non structured clouds that I
did not consider when I wrote that class.
Hi,

often organized clouds contain NaN values, e.g. when using Kinect data which
has "holes" where no depth value is measured.
Maybe those cause problems in your case?

Cheers,
Andreas


_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users


_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users


_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users