Skip to content

OmniCal MultiPose Alignment

One of the big frustrations of our powerful 3D workflow has always been achieving accurate 3D models and calibrating those - solutions so far have been based around laser scanning or having a skilled CAD person on site who can modify the CAD model when inconsistencies are found between the 3D mesh and reality.

OmniCal removes the requirement to have accurate 3D models using its powerful calibration and Mesh Deform tools.

Features

  • OmniCal uses structured light patterns to calibrate the relationship between projection surfaces, projectors and cameras.

  • The captured images are used to construct a 3D representation of projection surfaces as a point cloud.

  • Users then use the QuickAlign tool to manually align projection surfaces in the Disguise project to match their real world positions and proportions. This is an “offline” process which doesn’t require access to the physical stage.

  • We provide a single click Mesh Deform tool which deforms a 3D Model to match the real world (using the point cloud data)

  • Single click recalibration is supported if only projectors have moved. If cameras or projection surfaces have moved then an operator will need to adjust the previous alignment using current camera images.

  • The Disguise simulator also allows you to simulate cameras, view their coverage and perform test calibrations to ensure that the system will perform as required on site.

  • Supports 360° projection environments.

  • Designed for calibration of surfaces and scenes with 3D depth.

One camera option is available:

  • OmniCal MV system: For fixed installs. Reliable, continuous Ethernet camera connection; choice of lenses for three different fields-of-view; cameras only need setting up once.

Current limitations

  • The quality of the calibration will depend on having suitable lighting conditions.

  • Requires constant light levels during capture process.

  • Works best with low ambient light levels.

  • As mentioned above a Simulation must be run first to check project suitability. For large projects we recommend doing separate calibrations with groups of projectors and manually blending overlaps between the groups.

  • Requires non-reflective, opaque projection surfaces (no gauzes or mirrors).

  • Needs a few clearly defined feature points on the 3D mesh and real object, which can be visually identified on the camera images. Projection surfaces with sharp corners work well for example, but NOT smooth surfaces with no features.

  • OmniCal requires depth from the projectors point of view. Scene depth is particularly important when using moving elements such as Automation.