In pose estimation cpp project, estimation is based on cropped image created by 2d-landmark positions, then in visualization we must use cropped image to project 3d model and rasterize it.
If I use full image then translation and scale is false. I have fixed the scale problem by multiplying 3d model with scale calculated in getCroppedIm method, but translation is not fixed.
What is the solution for full image visualization?
Thank you.
In pose estimation cpp project, estimation is based on cropped image created by 2d-landmark positions, then in visualization we must use cropped image to project 3d model and rasterize it. If I use full image then translation and scale is false. I have fixed the scale problem by multiplying 3d model with scale calculated in getCroppedIm method, but translation is not fixed. What is the solution for full image visualization? Thank you.