I am trying to get the height of a flying object from the ground using single camera

I am trying to find the height of an object from the ground using single camera . Camera is place at top . Is there any way to get the distance of flying object to the camera from the 2d images.

Answer 1

[Bjorn Gustavsson]

Perhaps, provided that you know the physical sizes of the object and the field-of-view of your camera then it might be possible.

HTH

[andhavarapu lokesh]

Thank you sir for reply, how field of view is used here is there any relation to the object moving in the filed of view

[Bjorn Gustavsson]

Draw two lines intersecting with the angle of your field-of-view, plot a line perpendicular to the line bisecting the two lines (inside the field-of-view area), the part of that line between points where it intersects the f-o-v-lines illustrates your image detector. Then mark the imag location of the imaged object on that line. If you know the real-world size of that object you get the distance to the camera by the power of equal triangles. If you draw this with pen and paper step by step with some thinking between the steps you should get the simple pin-hole projection geometry of this task. This is obviously simplified to a 2-D world and 1-D image but the generalization to 3-D world and 2-D image follows naturally.

HTH

[andhavarapu lokesh]

sir actually i tried doing so , object near to optical axis is giving good result,but as the camera seeing the object at an angle then the values are not good. Can you give some suggestion how to correct.

Answer 2

[Jim Riggs]

There are two reasons why an object off-boresight looks smaller than an object that is in the center of the field of view.

Take an object which is in the center of the camera field of view (position (1)) and move it in the same plane to an off-boresight position (position (2)). The object at position (2) is now actually further away from the camera, and appears to be rotated by the viewing angle. Both of these effects cause the object to appear smaller.

[Image Analyst]

A third reason could be distortion (barrel or pincushion) by the lens.

[Bjorn Gustavsson]

No. That is the variation of solid angle as seen from the lens of an object translated from the optical axis. For the simplest pin-hole optics the image size of the two line-segments would be identical. For physical lens-systems with different optical projection functions (I dislike the sorting barrel and pincushion effects into the distortions) things become more complicated. What might be a more relevant problem is the proper shape of the object, see the modified image: