solar banner

What is SolAR?

SolAR is an open-source framework under Apache v2 license dedicated to Augmented Reality.


SolAR is modular and evolutive. It allows, in various contexts of use, the construction of customized computer vision pipelines targeting Augmented Reality applications (e.g. camera pose estimation or 3D mapping).


SolAR contains:

  • interfaces promoting interoperability
  • computer vision components constructed from third-party bricks
  • plugins for third-party applications

What's new in SolAR framework?
(Release notes)

v0.11.0 (2022/11/15)

  • Support of QR Code for relocalization
  • Support of PNG and JPEG decoding
  • New Front End service (for cloud deployment)
  • New Relocalization Markers pipeline and service
  • Optimization of Map Update, Mapping and Relocalization pipelines
  • CUDA versions of Map Update, Mapping and Relocalization services
  • Bug fixes, optimizations

v0.10.0 (2021/12/30)

  • Stereo camera components (e.g. calibration, rectification, depth estimation, reprojection to 3D)
  • Map update pipeline
  • SLAM and mapping optimizations
  • gRPC module: stubs for client and server sides, for all SolAR interfaces, to create remote services
  • Map Update, Relocalization and Mapping services (for cloud deployment) and client test applications
  • Add module PopSift for CUDA optimized SIFT feature detection and extraction.

SolAR framework Overview

An overview of our SolAR framework, an open-source solution dedicated to build Augmented Reality applications that offers cloud services.

SolAR ARCloud Services

First step - map initialization

This video shows the map creation process, using the 'SolaR ARCloud MapInit Service'.

Wearing a Hololens 2 headset, an operator is able to create the first part of the future global map. A fiducial marker is used only once, during this first initialization of the map, to define the 'World reference point' of the map.

Second step - map extension

In this video, an operator launches the 'SolaR ARCloud MapExtend Service' from an already mapped area and can explore a new area to extend the existing map.

Once the MapExtend Service has located the user in a previously mapped area, a corresponding portion of the global map is sent to the device. Then, the service is able to extend the extracted map according to the user's movement. The merging of this new extended map with the global map is done later in the background.

Relocalization service

This video presents the positioning service 'SolAR ARCloud Reloc Service' which is a service to quickly locate the user on the global map and track his movements.

Who needs SolAR?

According to your needs, SolAR framework offers several user profiles: