[Robotics] Project I wish to do for Blender for better Robot Navigation

[Jeremy NICOLA]

Dear Shreedhar,
if you are interested in using Blender for robot simulation, I strongly
advise you to check the Blender MORSE project:
http://www.openrobots.org/morse/doc/latest/morse.html .
https://github.com/laas/morse/

Here they are doing stereo vision with simulated camera:
http://vimeo.com/22246759#at=0

The development of the project and its mailing list are very active, so you
should definitely talk about your project on the mailing list:
https://sympa.laas.fr/sympa/subscribe/morse-users

Greetings,

Jeremy


2013/5/1 Shreedhar Pawar <shreedhar22 at gmail.com>

> Hello Sir,
>    I know that there are many of them who have written algorithms for
> depth estimation, but maybe the algorithms that I would be using, give more
> speed ups... like i would be using a parallel Boruvka's MST Algorithm for
> doing graph based image sementation, image  matching  in a parallel manner
> which would benefit from the segmentation done, that are the part of the
> depth estimation steps.
>    I hope I am putting my mails at the right place. Is there any special
> email id to put the proposals or is this the one...?
> Thanks for the advises.
> Shreedhar Pawar.
>
>
> On Wed, May 1, 2013 at 3:03 PM, Aris Synodinos <asynodin at mech.upatras.gr>wrote:
>
>> Dear Shreedhar,
>> Although I believe that your contribution in blender would be very
>> helpful, I would recommend to you to use Gazebo and ROS for such matters,
>> since the robotics community there is much more developed and more users
>> would be able to profit from your contribution in such matters.
>> Others have already implemented navigation/motion planning algorithms in
>> VGAs (CUDA and OpenCL) in various algorithms so I would suggest a proper
>> state of the art before starting to see what is already out there.
>> Best regards and good luck in your project,
>>
>>
>> On Wed, May 1, 2013 at 12:18 PM, Shreedhar Pawar <shreedhar22 at gmail.com>wrote:
>>
>>> Respected Sir,
>>>       I am interested in building a parallel algorithm for estimating
>>> depth of an image using stereo images on the OpenCL platform. Image depth
>>> estimation is really useful in robot navigation or in other areas of
>>> blender development like in gaming or 2D-3D video conversion.
>>>      There are many people who have build algorithms of estimating image
>>> depths, but the area of parallel computing with OpenCL is a bit unexplored.
>>> OpenCL is better than CUDA in  terms of portability(i.e. the code can be
>>> used with any graphic card or any multiprocessing CPU and even on handheld
>>> devices) which is the reason I choose to use it.
>>>      Please let me know whether I can do this project for Blender, so
>>> that I will send the details of the parallel algorithm and how it is better
>>> than any of the algorithms outside.
>>>
>>> Thanking You,
>>>
>>>
>>>   Your faithful,
>>> Shreedhar Pawar.
>>>
>>> _______________________________________________
>>> Robotics mailing list
>>> Robotics at blender.org
>>> http://lists.blender.org/mailman/listinfo/robotics
>>>
>>>
>>
>>
>> --
>> Aris Synodinos
>> PhD Student
>> Robotics Group
>> Mechanical and Aeronautics Engineering Department
>> Tel. +30-2610-996248
>> e-mail: asynodin at mech.upatras.gr
>> Web: http://robotics.mech.upatras.gr/
>>
>> _______________________________________________
>> Robotics mailing list
>> Robotics at blender.org
>> http://lists.blender.org/mailman/listinfo/robotics
>>
>>
>
> _______________________________________________
> Robotics mailing list
> Robotics at blender.org
> http://lists.blender.org/mailman/listinfo/robotics
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/robotics/attachments/20130501/c9c838c5/attachment.htm