Question

我使用openCV的stereoRectify时遇到很多问题，这些是我的变量：

    Mat K_02 = (Mat_<double>(3,3) << 9.597910e+02, 0.000000e+00, 6.960217e+02,
                     0.000000e+00, 9.569251e+02, 2.241806e+02,
                     0.000000e+00, 0.000000e+00, 1.000000e+00);

    Mat K_03 = (Mat_<double>(3,3) << 9.037596e+02, 0.000000e+00, 6.957519e+02,
                     0.000000e+00, 9.019653e+02, 2.242509e+02,
                     0.000000e+00, 0.000000e+00, 1.000000e+00);

    Mat D_02 = (Mat_<double>(1,5) << -3.691481e-01, 1.968681e-01, 1.353473e-03, 5.677587e-04, -6.770705e-02);
    Mat D_03 = (Mat_<double>(1,5) << -3.639558e-01, 1.788651e-01, 6.029694e-04, -3.922424e-04, -5.382460e-02);

     Mat R_02 = (Mat_<double>(3,3) << 9.999758e-01, -5.267463e-03, -4.552439e-03,
                     5.251945e-03, 9.999804e-01, -3.413835e-03,
                     4.570332e-03, 3.389843e-03, 9.999838e-01);
    transpose(R_02,R_02);

    Mat R_03 = (Mat_<double>(3,3) << 9.995599e-01, 1.699522e-02, -2.431313e-02,
                     -1.704422e-02, 9.998531e-01, -1.809756e-03,
                     2.427880e-02, 2.223358e-03, 9.997028e-01);

    Mat T_02 = (Mat_<double>(1,3) << 5.956621e-02, 2.900141e-04, 2.577209e-03);
    Mat T_03 = (Mat_<double>(1,3) << -4.731050e-01, 5.551470e-03, -5.250882e-03);

    Mat rotation = cv::Mat::zeros(3, 3, R_03.type());
    Mat translation = cv::Mat::zeros(1, 3, T_03.type());
    rotation = R_02 * R_03;
    translation = T_02 - T_03;

这些变量是使用KITTI数据集获得的，现在我想获得Q矩阵来重建场景，问题是使用stereoRectify：

Mat R1,R2,P1,P2,Q;
stereoRectify(K_02, D_02, K_03, D_03, size, rotation, translation, R1, R2, P1, P2, Q);

给我一个可怕的断言：

OpenCV Error: Assertion failed ((D.rows == ((flags & CV_GEMM_A_T) == 0 ? A.rows : A.cols)) && (D.cols == ((flags & CV_GEMM_B_T) == 0 ? B.cols : B.rows)) && D.type() == A.type()) in cvGEMM, file /build/buildd/opencv-2.4.9+dfsg/modules/core/src/matmul.cpp, line 3150

我无法调试它，因为gdb（通过Kdevelop）没有找到openCV的源代码（我已经使用存储库安装了它）。我在网上看到如何使用这个函数，但它总是使用stereoCalibrate函数的输出，所以我不知道哪个是用于定义我的内在函数矩阵的正确类型。

有人可以给我一个如何避免断言的提示吗？

Answer 1

问题似乎在矩阵translation中。您有错误，因为您试图将rotation（3x3）与translation（1x3）相乘。

只需转置translation matrix使用：

transpose(translation, translation);

或创建T_02和T_03而不是Mat_<double>(3, 1)

所以你有一个乘法Mat_<double>(1, 3)

cvGEMM在openCV中断言

1 个答案: