3D Pose Estimation
by Andrew Kirillov
The article describes application of POSIT algorithm for object's 3D pose estimation.
Introduction
자세추정은 많은 응용분야를 가진다. 3D pose estimation of an object from its image plays important role in many different applications, like calibration, cartography, object recognition/tracking and, of course, augmented reality.
많은 연구가 진행 되었다. There are number of research papers published about 3D pose estimation describing different algorithms.
가장 유면한건 POSIT알고리즘 이다. The most popular of them seems to be POSIT algorithm, which is quite easy to follow and is implemented in number of software libraries.
POSIT algorithm
약어 설명 : POSIT stands for POS with ITerations, where POS stands for Pose from Orthography and Scaling.
제안 논문 : The algorithm is described in "Model-Based Object Pose in 25 Lines of Code" paper by Daniel F. DeMenthon and Larry S. Davis.
OpenCV에도 구현물 있음 Implementation of this algorithm can be found in OpenCV library, for example.
목적 : So what does POSIT do? It estimates 3D pose of an object,
- which includes rotation over X/Y/Z axes and
- shift along X/Y/Z axes.
요구 사항 What does POSIT require to be able to do 3D pose estimation?
- First it requires image coordinates of some object's points (minimum 4 points).
- Very important to note is that these points must not be coplanar(동일평면) - i.e. they must not be all on the same plane.
- Then we need to know model coordinates of these points.
- This assumes that the model of the object we are estimating pose for is known,
- so we know coordinates of the corresponding points in the model.
- And finally the algorithm requires effective focal length of the camera used to picture the object.