AI - Classification

Original LIDAR

Fully automatic separation of non-ground from ground points

Automatic detection and classification of LIDAR points into buildings, trees, vegetation, traffic signs, light poles and many other objects through our deep learning capabilities.

  • Fast LIDAR processing & Classification – 18 GB approx 9 hours by regular PC

  • Highly accurate (+/- 5-10 cm) mapping & 80-90% accuracy of Classification

  • Compressed data model size – 1/10th of regular size

  • Rich experience in mapping with LIDAR and imagery

  • Cool add-ons to make LIDAR data processing compelling

  • Patented classifications

  • Highly effective automation process

  • Rapid processing of LIDAR (Airborne, Mobile, etc.) data into 3d imagery

  • Visionary, market oriented and driven


Collecting large amounts of data is important for both testing and situation analysis as well as supplying the training data needs of deep learning. YADO-VR and the way it has been build, complimented with powerful algorithms, is able to process large amounts of LIDAR generated point cloud data. The data collection is part of the Yado-VR base platform.


The Deep Learning capabilities of YADO-VR fuses all the essential data to accelerate the training and verification of the data gathered from sensor technology.. The Yado-VR ecosystem has all the functions to provide an automated DETECTION and automated CLASSIFICATION environment, making sure you end up with usable computer vision output to make the right decisions.


Relevant objects on the collected data have to be labeled precisely in order for the AI platform to learn all the classes reliably. The annotation and classifications functions  makes multisensor data annotation many times faster than the manual method.