Scan Result
Last updated
Last updated
After each scan, an intermediate step is displayed that shows the result of the projector scan. This step differs depending on whether ‘Flat Screen’, ‘Any Surface’ or ‘Curved Screen/Dome’ has been selected.
Here, the result of the scan is displayed as a generic test image on the projector and in a live feed from the camera's point of view. The scanning process is deliberately kept simple, so there are no additional parameters - everything is determined as automatically as possible.
The purpose of this measure is to assess whether the scan was successful. This is the case if:
The test image completely fills the projection image on the desired surface
no obvious and unexpected distortions occur
Please note that this step does not represent content mapping, i.e. the alignment of lines and structures does not play a role here.
Examine the scan if it meets the mentioned criteria. A good scan typically looks like this:
Click Next if everything is OK to proceed to the next projector.
If the scan result is not satisfactory, a new scan is necessary.
Here we see distortions of the projection image that is not caused by the surface and therefore represents an error in the scan:
Click Back to return to the Scan Settings and take into consideration:
Optimisation of the environmental conditions (light, obstacles, other factors)
Adjust the camera settings (focus, exposure)
Adjust the scan parameters: Dotsize, Threshold
After adjusting continue forward (Next....) to run the scan anew.
After each scan, an intermediate result is shown, representing the result of the data acquis
This step appears if the ‘Curved Screen/Dome’ scanning method has been selected (although this scanning method is also very suitable for other surfaces).
A good scan is represented by a full covered, distortion fee test pattern across the projector (minus masks is set prior to the calibration.
In addition, there is a visual representation of the scan in the Scan Inspector.
The visualisation of the scan result on the right-hand side is the most important information: the points captured by the camera are shown at the top of the camera image, with the resulting image mapping on the projector below
White dots represent positively recorded sample values, while red dots represent discarded samples. Black areas are not represented at all - either because these areas have no projection or because the detection of sample values does not work there.
A bad scan can be evaluated visually like described above for flat screen/any surface scan: obvious distortions, omitted projector coverage etc.
In addition, the scan inspector provides a visual feedback both of the scan dots and resulting projector coverage:
The camera picture shows few white dots, and red dots and vast areas of the projection area without any samples. This indicates a severe issue with the camera, e.g. bad signal, network issues, wrong settings, etc.
The projector coverage diagram below shows the result of the scan as it is displayed on the projector. Much space on the projector is left blank, so they are omitted for any coverage. The rest is computed but it is likely that due to the improper scan heavy distortions are happening.
The acceptance of sample points (white = OK, red = rejected) is controlled by the ‘Detection of invalid points’ parameter, among others, and is one of the most important parameters for analysing the scan.
The geometric scan is a rather complex procedure with many mathematical parameters to handle. We try to simplify the handling by determine a most useful setting automatically, so in most cases (= good camera conditions) not special treatment of these parameters is required.
However, the more demanding projection scenarios get, the more a fine-tuned scan adjustment pays off.
Here we describe the available parameters and their function. Please bear in mind that these parameters can be left untouched for most cases.
Geometry Scan Adjustment
Detection of possibly invalid points
How strict the algorithm is in the decision of valid/rejected points (blue vs red)
Extrapolation method
The software is able to extrapolate to regions of the screen where the camera was not able to scan.
Full polynomial means that every scanned dot should follow a homogeneous curve, e.g domes, curved screens.
Partial Polynomial means that the surface can be made out of different curvatures, non linear, e.g flat with round corners
Extrapolation distance
The more dots are chosen, the more the software tries to extrapolate. If you choose “Whole Display”, the software will try to extrapolate to the full display space. In case of a bad scan, the extrapolation might cause artifacts like swirls and deformed areas. The region of interest is an important aspect, if a projector is covering only a part of a screen, there is no need to extrapolate more than that area, therefore increment the extrapolation distance until you fully cover it
Surface curvature
If your surface has an extraordinary curve (slight or strong) you can get better results from switching between these options.
Intense overshooting of projector
Optimization of setups with an intense overshooting of the projectors
Optimization of Mesh
Pre-configured optimization of the texture distribution, enabled by default. Turn off if you notice strange patterns.
Crop Mask
Test image
Change the displayed test pattern with a custom image, e.g. load a 4k/6k grid to view higher details in the results and spot possible issues.
Save scan data
Save the result of this projector geometry scan in a file (.bdi). Use for backing up, comparing results or load during re-calibration.
Options
Test image
Change the displayed test pattern with a custom image, e.g. load a 4k/6k grid to view higher details in the results and spot possible issues.
Save scan data
Save the result of this projector geometry scan in a file (.bdi). Use for backing up, comparing results or load during re-calibration.
In VIOSO 6, the terms Full Polynomial Extrapolation and Partial Polynomial Extrapolation refer to how the calibration system extends the display area outside of the measured projector image, especially useful when warping and blending images for projection mapping.
Here’s a breakdown of the differences between the two:
Extends the calibration surface completely beyond the measured area using a mathematical model (polynomial). Use when you need to project beyond the camera-captured area — for example, projecting onto physical surfaces that the calibration camera couldn’t fully see.
Benefits:
Smooth continuation of the warp surface.
Ideal for projections onto known geometries (e.g. domes or curved walls).
Limits:
Can introduce distortion if the extrapolation goes too far beyond the actual measurement
Not useful if the surface is too complex or unpredictable.
Applies extrapolation only to a limited degree, often just to the outer edges of the measured region. Use when minor edge extension is needed — for example, to fill in gaps between projectors or to slightly extend the image for better coverage.
Benefits:
More flexible than full polynominal extrapolation.
Handles complex projection surfaces
Limits:
Produces easily slight errors like waves
Can also cause more heavy distotion if scans are bad recognized
An image file (.png or .bmp) that crops the geometry scanned and removes it entirely from the result. Example: you can load a projector mask with cropped borders to reduce the blending area size. Crop masks are created before scanning. Read more here:
