Tutorial on Quick and Easy Model Fitting Using the SLoM Framework

Christoph Hertzberg, René Wagner, and Udo Frese

Abstract

In many areas of experimental science ranging from robotics to psychophysical research, to evaluation of spatial sensor-data and surveying, model fitting is a ubiquitous subproblem. Often it is not the actual scientific goal but rather the necessary evil of calibrating the equipment. This tutorial introduces methodology and a library allowing to solve model fitting problems easily without requiring the user to have an in-depth understanding of this subject.

After a brief introduction to the theoretical background we guide the reader through using all main features of the SLoM C++ framework based on a stereo camera and inertial measurement unit (IMU) calibration example which is solved with less than 70 lines of non-problem specific code, and provide hints on applying SLoM to other classes of problems.

The reader is only assumed to have a working knowledge of C++ and a basic understanding of statistics and 3D geometry.

Paper

The paper as submitted to Spatial Cognition 2012 is available here (PDF).

Source Code and Data Set

Coming soon