FindEssentialMat for 2 drone images

,
1

Hello,

I have 2 images made by the calibrated drone. So I have all camera parameters.
Manually I select the same points on both images and pass this information to the function.

            CameraMatrix1[0, 0] = f;
            CameraMatrix1[0, 1] = 0.0;
            CameraMatrix1[0, 2] = cx;
            CameraMatrix1[1, 0] = 0.0;
            CameraMatrix1[1, 1] = f * ar;
            CameraMatrix1[1, 2] = cy;
            CameraMatrix1[2, 0] = 0.0;
            CameraMatrix1[2, 1] = 0.0;
            CameraMatrix1[2, 2] = 1.0;


EMatrix = CvInvoke.FindEssentialMat(ImagePoints1, ImagePoints2, CameraMatrix1);

Then I try to process the Essential matrix to receive the Projection matrix.

Matrix<double> D = new Matrix<double>(3, 3);
            Matrix<double> U = new Matrix<double>(3, 3);
            Matrix<double> Vt = new Matrix<double>(3, 3);
            CvInvoke.SVDecomp(EMatrix, D, U, Vt,0);

            if (U.Det<0) {
                U *= -1.0;

            }

            if (Vt.Det < 0) {
                Vt *= -1.0;
            }

            Matrix<double> W = new Matrix<double>(3, 3); //<< 0, 1, 0,
                                                         //  -1, 0, 0,
                                                         //   0, 0, 1);

            W[0, 0] = 0.0;
            W[0, 1] = 1.0;
            W[0, 2] = 0.0;
            W[1, 0] = -1.0;
            W[1, 1] = 0.0;
            W[1, 2] = 0.0;
            W[2, 0] = 0.0;
            W[2, 1] = 0.0;
            W[2, 2] = 1.0;

            Matrix<double> Wt = new Matrix<double>(3, 3); //<< 0, -1, 0,
                                                          //    1, 0, 0,
                                                          //    0, 0, 1);

            Wt[0, 0] = 0.0;
            Wt[0, 1] = -1.0;
            Wt[0, 2] = 0.0;
            Wt[1, 0] = 1.0;
            Wt[1, 1] = 0.0;
            Wt[1, 2] = 0.0;
            Wt[2, 0] = 0.0;
            Wt[2, 1] = 0.0;
            Wt[2, 2] = 1.0;

           

            R1 = U * W * Vt;
            R2 = U * Wt * Vt;

            t = U.GetCol(2) * 1.0;

            Matrix<double> P1n = new Matrix<double>(3, 4); //projection matrices in the new (rectified) coordinate systems for Camera 1.
            Matrix<double> P2s1 = new Matrix<double>(3, 4);
            Matrix<double> P2s2 = new Matrix<double>(3, 4);
            Matrix<double> P2s3 = new Matrix<double>(3, 4);
            Matrix<double> P2s4 = new Matrix<double>(3, 4);

CvInvoke.HConcat(CameraMatrix1 * R1, CameraMatrix1 * t, P2s1);
            CvInvoke.HConcat(CameraMatrix1 * R1, -1*CameraMatrix1 * t, P2s2);
            CvInvoke.HConcat(CameraMatrix1 * R2, CameraMatrix1 * t, P2s3);
            CvInvoke.HConcat(CameraMatrix1 * R2, -1 * CameraMatrix1 * t, P2s4);

But the result is completely useless. Please tell me - should I use only points on the flat surface for FindEssentialMat? What can be the reason for the wrong Essential Matrix?