Scene vs. Object Based 3D Laser Scanning

For years I have been struggling with how best to categorize the difference between the short range, metrology world and the “longer” range, <you fill in the blank> world. Although it would seem to be all about scale, it really is not that simple. In some cases the short range pros scan large items, like airplanes and the long range pros need to get fine detail as in a building facade.

Thanks to Karl Matthews, VP of Product Management at Geomagic I now have a simple, but elegant (those are the best kind) solution to this organizational problem.

Taxonomy is the science or technique of classification. That is what has been missing for me when I look at the two worlds of laser scanning. Without classification you don’t have the ability to organize your thinking. As an engineer I don’t like that. I need a taxonomy for the laser scanning worlds and thanks to Karl I think I now have it.

As he pointed out to me last week, the best way to think about the 2 worlds of laser scanning is not in terms of scale, but rather in terms of objects and scenes. When scanning an object your point of view is inward, toward the object. When scanning a scene your point of view is outward, with the scanner at the center of the point cloud.

Beautiful, thanks Karl – problem solved, as far as I am concerned, but not so fast. Karl does not see it quite the same way.

The reason Karl developed this taxonomy was to organize his thinking about software. Turns out object centric software is quite different from scene centric. Geomagic is very good at object centric point cloud processing. They have not really spent much time on scene centric point cloud algorithms.

I have a feeling that could change. Stay tuned.

 

This entry was posted in Education, Metrology, The Industry. Bookmark the permalink.

2 Responses to Scene vs. Object Based 3D Laser Scanning

  1. James says:

    Now there’s an epiphany I can sink my teeth into!

    Thanks for sharing Gene.

  2. This is a way to look at it, I’ve used these terms as well, but its not that simple. We frequently use ‘long range’ or ‘scene’ scanners (in this taxonomy) for object style scanning. Examples such as a car, monument, or airplane are typically described as objects, not scenes.

    I do agree that despite the use, the various software products usually specialize in one or the other type of post-processing: object or scene. It would be great if more software could do both styles from more scanners, long or short range, for either object or scene-based use.

    Perhaps we’ll see new solutions next week at SPAR. I always find it very interesting as to which software firms attend to learn more about these markets and uses.

Leave a Reply

Your email address will not be published. Required fields are marked *


*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>