MonocularSLAMBaxter

Contents

• Introduction
• Usage
• Result

Introduction

A naive algorithm to solve monocular SLAM problem in C++ using Baxter robot.

Usage

1. Enable the Baxter robot and open its camera.

rosrun baxter_tools enable_robot.py -e
rosrun baxter_interface joint_trajectory_action_server.py --mode position
rosrun baxter_tools camera_control.py -o left_hand_camera -r 1280x800

2. Save the pictures.

rosrun MonocularSLAMBaxter vision.py

3. Let the robot move.

rosrun MonocularSLAMBaxter move.py

4. Process the data collected.

Compile and execute dense_mapping.cpp .

Result

Feature extraction and matching.

The depth image after 5 iterations.

The depth image after 100 iterations.