Difference between VertexCam and VertexSE3

[haoruij677]

Hi,

I'm wondering what's the difference between VertexCam and VertexSE3 and if I can convert one type to another. The reason for this question is that I would like to use EdgeSE3PointXYZ to optimize the pose according to the transformation from 3D world coordinates to 3D camera coordinates. But what I have in hand is just VertexCam instead of VertexSE3 type, therefore my thinking cannot be implemented. Is there any way to resolve this problem or is there another edge like this defined on VertexCam and VertexBAPointXYZ type? Thank you so much~

[haoruij677]

Sorry I was not making myself very clear. I'm currently using EdgeProjectP2SC, which is defined on VertexCam and VertexBAPointXYZ, to do bundle adjustment. Meanwhile, I would like to add another edge constraint on my pose, which aims to optimize the pose according to the transformation from 3D world coordinates to 3D camera coordinates. I was thinking about using EdgeSE3PointXYZ to deal with this but it turns out that this edge only works for VertexSE3 type vertices. Therefore, I'm asking if there's a way to deal with this kind of situation. My thinking is we can either construct another VertexSE3 pose vertex and keep it the same as the VertexCam and optimize these two together, or is there another predefined edge that can solve this issue and is defined on VertexCam and VertexBAPointXYZ type? Thank you so much~