Feature Extraction & Segmentation

Summary

Technology Description

A common task in systems that utilize LiDAR data is the extraction of useful features from the cloud of points that LiDAR systems produce. This task is vital, and Norvana offers an efficient solution to the problem of feature extraction and segmentation of point cloud data. Given a collection of LiDAR data with 1 million points Norvana can process it within a second. Furthermore, the Norvana system is autonomous. Many applications involving LiDAR currently require immense amounts of human power, this technology allows these processes to be automated. This has benefits for any technology that involves extensive manipulation of LiDAR data. The global LiDAR market is currently a billion-dollar industry and is poised to increase as the advent of self-driving cars has given rise to a new demand for LiDAR and surrounding technologies. The Norvana system in its current form has been tested on mobile and terrestrial systems. This means that it is applicable to a significant portion of the global market. Norvana could see use in geospatial projects, various types of mapping projects, and construction. Norvana could also be extended far beyond the presently tested applications and see use in any application that requires feature extraction and segmentation. This is of great benefit because feature extraction and segmentation are fundamental parts of all LiDAR based applications and with Norvana you now longer need to choose between speed and accuracy.

Features & Benefits

  • Processing 1-2 million data points within a second
  • Completely automates aspects of LiDAR processing that are otherwise time consuming
  • Experimentally confirmed to be more robust and accurate than competing products

Applications

  • Geospatial
  • Mapping
  • Construction
  • Airborne
  • Autonomous vehicles
  • Augmented Reality

Background of Invention

This technology was produced over the course of three years of research by the School of Civil and Construction Engineering at Oregon State University by Michael Olsen, Associate Professor of Geomatics and Editor-in-Chief for the ASCE Journal of Surveying Engineering; and Erzhuo Che, postdoc at OSU. Development was supported by the National Science Foundation and Oregon Department of Transportation (ODOT).

Status

U.S. patent pending; available for licensing.

   

 

Patent Information:
Tech ID:
OSU-19-48
Category(s):
Software
Transportation
Contact:
John Sweet
IP & Licensing Manager
Oregon State University
john.sweet@oregonstate.edu
Inventors:
Erzhuo (Ezra) Che
Michael Olsen
Keywords:
lidar
© 2020. All Rights Reserved. Powered by Inteum