Q1: Why do we need to pose estimation?
from document referring to “how the object is situated in space, like how it is rotated”, what means to know how an object place in 3D space, but render on 2D plane image, the core idea is converting a 3D point to a 2D pixel point it is a project from 3D coordinate to 2D coordinate
Q2: How to estimate?
convert the problem to “where the camera position in 3D space, if shot the image(chessboard) vertical in Z, and parallel in XY plane” then from the document said “we can assume Z=0, such that, the problem now becomes how the camera is placed in space to see our pattern image.”
Q3: How does it work?
- we just prepare some vertex
- find rotate & transform vectors from camera matrix & distort coefficients
- then project vertex to pixel use vectors
- draw those vertexes
