Active Structure-from-Motion for 3D Straight Lines

TL;DR

This paper develops an active structure-from-motion framework specifically for 3D straight lines, enabling reliable estimation of their parameters for applications like visual servoing, validated through simulation and real robot experiments.

Contribution

It introduces an explicit representation and change of variables for 3D straight lines, extending existing frameworks to handle these features for improved robustness.

Findings

Framework successfully estimates 3D line parameters in simulation.

Real robot experiments confirm effectiveness for single and multiple lines.

Minimum of three lines needed for full 6-DOF camera control.

Abstract

A reliable estimation of 3D parameters is a must for several applications like planning and control. Included in the latter is the Image-Based Visual Servoing, whose control scheme depends directly on 3D parameters e.g. depth of points, and depth and direction of 3D straight lines. Recently a framework for Active Structure-from-Motion was proposed, addressing the former feature type. However, straight lines were not addressed. These are 1D objects, which allow for more robust detection and tracking. In this work, the problem of Active Structure-from-Motion for 3D straight lines is addressed. An explicit representation of this type of feature is presented, and a change of variables is proposed, which allows the dynamics of the line to respect the conditions for observability of the framework. The approach is validated first in simulation for a single line, and second using a real robot. The latter set of experiments are conducted first for a single line, and then for three lines, which is the minimum required number of lines to control a 6 degree of freedom camera.