Skip to content

tek5030/lab-pose-estimation

Repository files navigation

Pose estimation and Augmented Reality

Welcome to this lab in the computer vision course TEK5030 at the University of Oslo.

In this lab we will implement pose estimation based on 3D-2D correspondences from scratch!

Illustration of pose estimation from 3D-2D correspondences

Start by cloning this repository on your machine.

Then, open the lab project in CLion using the cmake-file in the base directory: lab-pose-estimation/CMakeLists.txt. If you are uncertain about how this is done, please take a look at the intro lab.

The lab is carried out by following these steps:

  1. Get an overview
  2. Camera calibration
  3. Implement homography-based pose estimation
  4. Implement Motion-only Bundle Adjustment

You will find our proposed solution at https://github.com/tek5030/solution-pose-estimation. Please try to solve the lab with help from others instead of just jumping straight to the solution ;)

Please start the lab by going to the first step.

Prerequisites

  • OpenCV must be installed on your system. If you are on a lab computer, you are all set.

    If you are on Ubuntu, but not on a lab computer, the following should be sufficient for this lab.

    sudo apt update
    sudo apt install libopencv-dev libopencv-viz-dev
  • We refer to setup_scripts and the intro lab as a general getting started-guide for the C++ labs on Ubuntu 22.04.