I have reviewed various methods for calibration and their calculations. I would like to confirm the precise differences between the following terms: Error allowance, standard error estimate, tolerance, and accuracy. Is there anyone available for a discussion on this topic? Thank you.
Look at a document called VIM
Look at the Publications section on BIPM website. Its under Guides in Metrology.
www.bipm.org
I am not a specialist, but I have implemented with success a camera calibration process (https://data.inpi.fr/brevets/FR2810830).
Error calculation sounds like an estimation, you may calculate this from the theoretical formula if you have one (delta...). In my case the result was better than the estimation because some errors are linked together, and also because a variables (like gaussian centered) are not "worst case" all at the same time, and this achieves some magical compensation.
My model was hybdrid, with some part theoretical (3D rotation from encoders, telemetry from laser) and some other parts with lookup tables manually built because model was too complex (zoom and focus). We had added some user interface (sliders) to tweak some parameters and try to better understand instabilities. Thus you could modifiy the variable by injecting manually somme error, and watch the result il real time.
Hope this helps.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films