Precision Measurement with Target Marks

ELCOVISION 10 can use targets to measure single, precisely defined points with maximum accuracy. While automatically measured natural points have a measurement accuracy of about 0.2 to 0.5 pixels in the image, targets exceed this with an impressive 10 to 100-fold increase in accuracy on average. This precision in image measurement results in equally accurately determined object points, whereby point accuracies in the sub-millimetre range can be achieved almost effortlessly.

The typical measurement accuracies behave as follows for different target mark diameters D:
Diameter D Centre circle | Typical accuracy with which the centre of the target mark is measured | Typical* point accuracy with a pixel size of 1mm on the object |
10 – 15 Pixel | 91% better than 0.10 Pixel 70% better than 0.05 Pixel 21% better than 0.02 Pixel 14% better than 0.01 Pixel | 91% better than 0.075mm 70% better than 0.037mm 21% better than 0.014mm 14% better than 0.007mm |
15 – 20 Pixel | 94% better than 0.10 Pixel 80% better than 0.05 Pixel 31% better than 0.02 Pixel 20% better than 0.01 Pixel | 95% better than 0.075mm 80% better than 0.037mm 31% better than 0.014mm 20% better than 0.007mm |
20 – 30 Pixel | 98% better than 0.10 Pixel 93% better than 0.05 Pixel 39% better than 0.02 Pixel 25% better than 0.01 Pixel | 98% better than 0.075mm 93% better than 0.037mm 39% better than 0.014mm 25% better than 0.007mm |
30 – 40 Pixel | 99% better than 0.10 Pixel 94% better than 0.05 Pixel 42% better than 0.02 Pixel 28% better than 0.01 Pixel | 99% better than 0.075mm 94% better than 0.037mm 42% better than 0.014mm 28% better than 0.007mm |
40 – 50 Pixel | 99% better than 0.10 Pixel 98% better than 0.05 Pixel 53% better than 0.02 Pixel 36% better than 0.01 Pixel | 99% better than 0.075mm 98% better than 0.037mm 53% better than 0.014mm 36% better than 0.007mm |
50 – 70 Pixel | 99% better than 0.10 Pixel | 99% better than 0.075mm |
*When using a higher quality camera/lens combination and good image quality, 4 or more image measurements per target mark and good image capture configuration.
If the images are made in such a way that a target mark is visible in at least 8 images, the few "outliers" with poor image measurement accuracy can be identified and deleted using the global orientation, object point accuracies of better than 0.05mm are relatively easy to achieve.
Application Examples for Target Markers:
ELCOVISION 10 recognises 2 versions of targets, V1 and V2:
Targets V1
- Suitable for relatively poor image quality such as noisy, slightly blurred, underexposed or overexposed image.
- 12 bit barcode: 47 different targets.
Targets V2
- Need good image quality.
- 16 bit barcode: 800 different targets.
![]() https://www.elcovision.com/Marker/TargetV1V2.zip
|
![]() |
![]() |
