3D Vision

Our core competencies

Our software solutions process, analyze and match depth maps and 3D data from 3D sensors like LiDAR, laser scanner, structured light, ToF cameras, stereo cameras, and many more.

From point clouds to 3D models with efficient, robust 3D algorithms

From depth maps to 3D models

Our high performance 3D matching algorithms merge thousands of individual 3D scans by SLAM to a consistent 3D model. Our versatile 3D matching algorithms (rigid, deforming, intersection-free) process all common geometric representations (point clouds, meshes, NURBS).

Advantages and benefits
Synthetic yet ultrarealistic human 3D models for machine learning

Synthetic ultra-realistic human 3D models for machine learning

Training and validation of machine learning algorithms on human faces demands for extensive and high-quality data sets. Acquiring such data is expensive, legally complex (privacy issues) and partially impossible (e.g. accidents in autonomous driving).

Our software methods for the generation of synthetic yet realistic 3D data of human faces and bodies address this challenge. From a small number of 3D scans of real persons, we generate a nearly infinite number of realistic new 3D models through “interpolation”. Together with a few real data they serve as data for training and validation of machine learning methods for example in in-cabin monitoring for automated driving.

Creating and analysing digital twins

Simulation on real world 3D data (such as digital elevation data)

A special application area of our high performance algorithms is the processing of terabyte sized 3D data sets like digital elevation models and 3D infrastructure scans. We develop and operate a high-performance web service for the Austrian federal states, providing solar altitude analyses, elevation requests and profile services with almost real-time responses to user queries. In the area of infrastructure our software methods translate raw point scans into high-precision digital twins respectively BIM models.

Advantages and benefits

Do you have any questions? We're looking forward to receiving your message.