[ros-users] small text for report which includes object detection

René Ladan r.c.ladan at gmail.com
Thu Aug 5 08:33:04 UTC 2010


2010/8/4 Dejan Pangercic <dejan.pangercic at gmail.com>:
> Hi Rene,
>
> again sorry for the late chime in.
>> •       Get an image of the object (e.g. a pool ball) from e.g. a stereo
>> camera with texturing or e.g. a laser range finder.
> I am not sure what are you trying to say with "camera with texturing",
> but what has been done is applying the following stereo processing
> code on the stereo images:
> https://code.ros.org/svn/ros-pkg/stacks/image_pipeline/trunk/stereo_image_proc/mainpage.dox
> Now, to improve the quality of the stereo correspondences, a pattern
> projector has been uses to boost up the number of corresponding images
> in both cameras
> (http://www.willowgarage.com/blog/2009/06/19/sensor-head). Shall one
> use a laser range finder, then it is important to notice that the
> finder has to be tilted in order to generated the full 3D point cloud.
>
What you just said above. The usual problem of taking a distance
from the text occurred.

>> •       Transform the image (either a stereo-photo or a “sphere” of
>> distances) into a point cloud. A point cloud is a collection of
>> points, where all points have a number of pre-defined attributes, like
>> spatial coordinates, color, or neighborhood density. Point clouds are
>> the universal data structures for performing 3-dimensional algorithms
>> within ROS. TODO: figure out the algorithm
> Do you still need more clarifications here?
>
Not for now, the report is still work in progress.

>> •       Transform the point cloud into a wireframe. TODO: figure out which
>> algorithm is used for this.
> Do you still need an example of this? I got some code but would need
> to make it pcl-compatible first.
No, I think Matei explained it quite well in this thread already.

Regards,
Rene



More information about the ros-users mailing list