Low Level Detection in Dense could of points


Growing use of Laser Scanner creates new needs among their users, in which software plays a critical role. Among these needs is the painful and cost productive task of retrieving low level information from dense point’s clouds, for instance main break lines or main components of a scanned structure. For such tasks, there are no satisfactory automated tools at our knowledge, and the job is almost done manually.

Such applications are a tempting perspective for the Laser Scanner industry: the added value is clear for the software business involved in the processing and engineering of dense points clouds, but also impacts Laser Scanner makers, that may focus more efficiently prospect clients primarily targeting those functionalities.

These problems are also very challenging and promising for the scientist community, since it requires a new approach that may be considered as a step towards results and applications in other applied sciences.

However, to set on a successful research and development program able to develop a software solution capable to address some of these new HDS Laser Scanner needs requires blending Laser Scanner industry experience and fundamental research capabilities.

CRIMERE invested in a research program, called "Optimally Transported Schemes", for which the representation of objects or measure by a finite number of points is a central thematic. For Laser Scanner applications, CRIMERE thinks now that the following applications can be targeted:

  • Automatic detection of break lines of a scanned object, in order to answer to clients needing a filar or planar representation of objects, as urbanisms architects.
  • Identification of main structure of scanned objects. It is more precisely a segmentation of dense clouds into zones sharing close information as refraction index or color codes.
  • Compression of dense clouds, targeting a compression algorithm to reduce drastically the cloud size while minimizing the loss of information (as .jpeg or .avi compression format for the image industry).

CRIMERE is trying to federate a research program around these thematics. We welcome any research institutes or industrial actors willing to contribute to these topics.

Your rating: None