Question about "ImageFileProcessor.cpp"

[zengletian1491]

In the file “3DUNDERWORLD-SLS-GPU_CPU/src/lib/core/ImageFileProcessor.cpp”, line 90-94: I think the expression of camTransMat_ has something wrong!

Refer to https://docs.opencv.org/2.4/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html

In my opinion, camTransMat_ = - R.transpose() * t, where R denotes the rotation matrix and t is the translation vector.

However, after the twice transpose operations of R in line 92 and line 93, the camTransMat_ in line 94 is “- R * t”， not “R.transpose() * t”.

Could you share your ideas about the aforementioned? Thanks!

[v3c70r]

Hi, I think it's because of the GLM library I used at that time. The matrices in in GLM are column major where the OpenCV's matrices are row major. The first loop is used to convert the row major matrix to column major matrix.
But you are absolutely right about the idea that I have implicitly transposed the matrix twice, which is unnecessary. There's no need to have another transpose(inverse) there in line 93 IF line 92 is assigned in the transposed order. I'll probably make a change on that.

[zengletian1491]

OK, thank you.
Line 111 and line 145 in fileReader.cpp should also be changed.