documentation

Version Info

Release date: 31.01.2020 Release info: General Release

General

This is a major update for VisionLib which adds:

Tracking of multiple models using the MultiModelTracker. Please refer to the Multi Model Tracking Tutorial.
Support for ARCore. Please refer to the external SLAM Tutorial.
Many bugfixes and enhancements.

Thank you to all our users for giving feedback.

Bugfixes

General

An issue has been fixed, which didn't release the camera when the application has been suspended.
An issue has been fixed, which used an init pose in AutoInit mode even if none has been defined.

Enhancements/Features

Camera Calibration: chessboard is now better detected on high resolution or blurry images.
Added MultiModelTracker for tracking several models simultaneously. MultiModelTrackingExample.unity has been added as an Unity example scene.

Platform specific changes

iOS

Resolved a problem, where device and screen orientation has not been respected, when using the objective-c interface. Small adaptions in objective-c tutorial and example scene have been made.

Android

Added support for ARCore. Please refer to the external SLAM Tutorial.

Windows

Added synchronous example for the native C-API. Images are acquired with OpenCV and injected into VisionLib.

Interfaces

Unity

The VisionLib example scenes have been renamed.
Added MultiModelTracking.unity as an example scene for the new multi model tracking feature.
Added visual output for tracking events.
Renamed the VisionLib example scenes.
An issue has been fixed that prevented camera calibrations to be performed on image sequences.

Documentation

Updated Licensing information regarding Multi Model Tracking (Beta).
Added the Multi Model Tracking Tutorial.
Added a page that shows how to work with VisionLib Tracking Events.
Updated the Using VisionLib page to give an overview of VisionLib workflows.
Added a News and Updates page to summarize updates concerning VisionLib.
Separated the AR Background from the Introduction page.