Ok solved it. first I gonna use a Matlab tool and the chessboardthingie to get the intrinsic parameters. And the i ll use cv::solvePnP()
Yeah I used the search but unforunately it didnt help me alot.
As the title says I have to find the camera-coordinates in real world. But I am also facing some problems concerning opencv commands. Till now I know some points in the image and the corresponding World coordinates. (worst case 6 points are known)
To get the camera-cooridnates I thought of using the formula:
transp(x,y,z)= - transp(R)*t , where R and t are the extrinsic parameters
so now my proble is: how to calculate R and t ? I mean does an opencv function exist that solves this problem? Cause actually there are a lot camer calibrate functions. Or do I have to first calculate the intrinsic parameters ?
edit: here is the function i meant: http://opencv.willowgarage.com/documentation/python/calib3d_camera_calibration_and_3d_reconstruction.html#calibratecamera2
but i dont know how the parameters should look like. How can I initialize manually a for example 3x6 Matrix like cv::Mat ([[x1] [y1] [z1]] .... [[x6] [y6] [z6]])
