Calibration Camera

[masrazie]

Hello Jane and dear All,

I am CalCam user from Indonesia,

just a little asking about the ' Template' of the camera calibration,

1. Due to Our Office space in field implementation it is enough accuration using templete size : 2 m x 1.5 m for SLR Camera type, for example: Canon 5D.

2. every camera type POCKET or SLR should be 'print' templete and change templete ?.

Regards,
Andreas

[Janne]

Andreas,

Theoretically you should use the calibration 'template', or wall size, that CalCam calculates for your camera-lens. For practically all SLR cameras the optimal wall size is quite big, for example 11 x 7 meters.

In practice, as it is many times not possible to establish the calibration targets of optimal size, you should set-up the targets as big as your office space allows. We have often used targets of about 3m x 2m for all cameras - the images come out somewhat unclear (because of focus in infinity) but the locations for the targets are correct and identifiable for the software. 2m x 1.5m sounds a bit small, especially if you have long focal length lens. You should try to increase the size to the limit that your office space allows.
If you cannot increase, you can try that size, calculate the parameters and analyze the residuals.

You can use the same targets for all cameras. Again, theoretically you should establish a different size field for all camera-lens -combinations, but in practice you can use the same field for all. It will serve you future cameras better if you increase its size to maximum allowable.

Janne

[Yau Yeng Yew]

Dear Janne,
Is it necessary to set the camera focus to infinity while taking calibration shots? & there after tape the focal ring so that it stays until the next calibration?

Yau

[Janne]

Hello Yau,

Yes, you have to set the focus to infinity. When taking calibration images you'll have to have the camera with exactly the same settings which you use for imaging. Apply the same aperture that you use in-flight.
The correct procedure is to fix the focal ring to infinity with tape. Some lenses allow to use a screw which is better than tape.

Janne

[dMk]

Dear,

I have take 10 images and did the calibration (Calcam 1.1.1) twice, both of the calibration results data show some deference.

Can you tell me why this difference occurs?
Did you have standard deviation on previous calibration results (on each particular camera)?
Did you ever compare the calibration result of the images taken by autofocus setting with calibration result of images taken by infinity focus setting (on the same camera)?

Best Regards
Karibo

[Janne]

When calibration results show some difference there are a few possible causes:
1. The calibration targets have moved between exposures (not firmly attached)
2. There are differences in editing the labels, differing residuals
3. The camera itself has been changed, i.e. lens or sensor are not in the same position.
It is normal to have differing results. It is something to worry about if they differ significantly, not only differing residuals.

We have not compared results between Autofocus setting and Infinity setting. You must take the calibration images using the same settings you will use during imaging. I.e. Infinity setting, even though calibration images may become a bit unclear because of out of focus.

[dMk]

Dear Jeane,

I meant, I did the calibration twice on the same 10 images I've take.
Both results difference one another.

Thanks.

[Janne]

Ok, now I understand.
Again, it is possible that the results differ, especially if
- you have edited points manually.
- any of the input values has been different
- constraints or initial values have been different, even slightly.
Don't worry too much if the results differ. Concentrate on the residuals and significance of the parameters.