A Backpack-Mounted Omnidirectional Camera with Off-the-Shelf Navigation Sensors for Mobile Terrestrial Mapping:...

Area of Science:

• Geomatics Engineering
• Robotics and Autonomous Systems
• Forestry and Environmental Science

Background:

• Personal Mobile Terrestrial Systems (PMTS) are increasingly used for mobile mapping in challenging terrains like forests.
• Emerging technologies like miniaturized navigation sensors and omnidirectional cameras enable low-cost PMTS.
• Existing off-the-shelf sensors often lack high-accuracy metric capabilities, necessitating advanced processing.

Purpose of the Study:

• To develop and evaluate a low-cost PMTS for forest mapping applications.
• To assess the performance of off-the-shelf navigation systems and omnidirectional cameras in a forest environment.
• To contribute to the advancement of lightweight, affordable PMTS for ecological and mapping purposes.

Main Methods:

• Integration of an omnidirectional camera with off-the-shelf navigation systems to form a low-cost PMTS.
• Implementation of an integrated sensor orientation approach utilizing navigation data for initial trajectory estimation.
• Data acquisition and processing within a forest environment to generate 3D point clouds.

Main Results:

• The integrated sensor orientation approach significantly improved trajectory accuracy, particularly in areas with limited satellite visibility.
• The generated point cloud achieved accuracy consistent with the Ground Sample Distance (GSD) of the omnidirectional images (3.5-7 cm).
• The PMTS demonstrated reliable performance comparable to other existing PMTS approaches in forest environments.

Conclusions:

• The proposed low-cost PMTS is a viable solution for mobile mapping in forest environments.
• Utilizing navigation data as initial information enhances trajectory accuracy in challenging conditions.
• Off-the-shelf sensors can be effectively employed for metric mapping applications with appropriate processing methods.