Abstract
Abstract. In this study we present a Micro Aerial Vehicle (MAV) equipped with precise position and attitude sensors that together with a pre-calibrated camera enables accurate corridor mapping. The design of the platform is based on widely available model components to which we integrate an open-source autopilot, customized mass-market camera and navigation sensors. We adapt the concepts of system calibration from larger mapping platforms to MAV and evaluate them practically for their achievable accuracy. We present case studies for accurate mapping without ground control points: first for a block configuration, later for a narrow corridor. We evaluate the mapping accuracy with respect to checkpoints and digital terrain model. We show that while it is possible to achieve pixel (3-5 cm) mapping accuracy in both cases, precise aerial position control is sufficient for block configuration, the precise position and attitude control is required for corridor mapping.
Highlights
Unmanned Aerial Vehicles (UAV) are nowadays a wellestablished tool for mapping purposes
The rootmean-square (RMS) of check point residuals is depicted in Table 1 and is 0.043 m horizontal and 0.040 m vertical, which corresponds in both directions approximately to 1 pixel
It should be noted that this accuracy is comparable with that of indirect sensor orientation using the current check points as ground control points
Summary
Unmanned Aerial Vehicles (UAV) are nowadays a wellestablished tool for mapping purposes. If we neglect the issue of safety, maintaining good geometry and assuring demanded accuracy of the final mapping products are the main difficulties The latter depend on the number and distribution of ground control points and/or accuracy of directly measured exterior orientations (EO) of imaging data. Several rather long corridor mapping projects have been presented using indirect sensor orientation approach (Delair-Tech, 2014; senseFly, 2015) These data collection activities require time- and labourintensive efforts which may be difficult to realize in corridors surrounded by high vegetation. The limited weight, volume and power availability in a MAV payload poses a major challenge when determining accurate orientation on-board This is especially true for inertial sensors rather than for GNSS receiver/antenna equipment.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
