Revision history [back]

Hi,

I was wondering – is it possible to achieve a decent approximation of the intrinsic parameters of a camera by using the fact taht you know the the current (optical) zoom of a camera? AFAIK the intrinsic parameters represent the principal point and the focal points in pixels. From my understanding, the current focal length can be calculated by multiplying the zoom level by the minimum focal length. After that one can calculate the focal length in pixels by using the following formula (horizontally in this case):

focal_pixel = (focal_mm / sensor_width_mm) * image_width_in_pixels

Analogically one can find out the vertical focal length. Further more, assuming that the principal point is through the middle of the screen and that the camera has 0 skew, would that result in a wrong intrinsic matrix?

Hi,

I was wondering – is it possible to achieve a decent approximation of the intrinsic parameters of a camera by using the fact taht you know the the current (optical) zoom of a camera? AFAIK the intrinsic parameters represent the principal point and the focal points in pixels. From my understanding, the current focal length can be calculated by multiplying the zoom level by the minimum focal length. After that one can calculate the focal length in pixels by using the following formula (horizontally in this case):

focal_pixel = (focal_mm / sensor_width_mm) * image_width_in_pixels

Analogically one can find out the vertical focal length. Further more, assuming that the principal point is through the middle of the screen and that the camera has 0 skew, would that result in a wrong intrinsic matrix?

Hi,

I was wondering – is it possible to achieve a decent approximation of the intrinsic parameters of a camera by using the fact taht you know the the current (optical) zoom of a camera? AFAIK the intrinsic parameters represent the principal point and the focal points in pixels. I.e.

focal_length = zoom_level * minimum_focal_length

From my understanding, the current focal length can be calculated by multiplying the zoom level by the minimum focal length. After that one can calculate the focal length in pixels by using the following formula (horizontally in this case):

focal_pixel = (focal_mm / sensor_width_mm) * image_width_in_pixels

Analogically one can find out the vertical focal length. Further more, assuming that the principal point is through the middle of the screen and that the camera has 0 skew, would that result in a wrong intrinsic matrix?