Evaluation in Real World of the Measuring Position Determination for Visual Inspection using UAV

Asa, Kotaro; Funabora, Yuki; Doki, Shinji; Doki, Kae

doi:10.1109/iecon.2018.8591270

Cited by 3 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, they did not indicate the exact poses that UAVs should collect data, which is indispensable for practical inspection. Contrarily, our CPP method [35], [36] divides the 3D model into as few clusters as possible and plans poses for each cluster. As a result, the UAV only needs to collect images at sensor poses and complete coverage can be achieved, the sensor poses can also be minimized.…”

Section: Related Workmentioning

confidence: 99%

“…The inspection target is presented as a set of triangular polygons M tar (STL [54] format) in virtual space V 3 . The M tar can be a set of planar or uneven surfaces, bridge piers, or dome structures of arch bridges and tunnels (see [35], [36] for more path samples). The size L and resolution I of the image sensor of UAV's camera in the direction of its short edge, and the focal length f are known.…”

Section: Automatic Surface Data Collection Using Uav a Problem Formul...mentioning

confidence: 99%

“…However, although the importance of data GSD has been recognized, the path planning method that ensures GSD has not been well studied. In previous research [35], [36], we have proposed a coverage path planning (CPP) method that uses targets' 3D models to plan their inspection paths. We validated it in simulations and confirmed that it can inspect targets under ideal conditions.…”

Section: Introductionmentioning

confidence: 99%

“…To further validate our previous algorithm [35], [36] and clarify if it needs enhancement, we implemented it to inspect some surfaces of a bridge in this study. Results indicate that the targets are almost successfully inspected and the GSD of collected data is qualified, yet the positioning errors of UAV resulted in a small percentage of omissions.…”

Section: Introductionmentioning

confidence: 99%

“…The section II reviews the works related to inspection path planning. The section III defines the inspection data collection problem we focus on and briefly introduces our CPP method [35], [36]. The section IV introduces the practical experiment and how we planned the flight paths for the UAV.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Evaluation and Enhancement of Resolution-Aware Coverage Path Planning Method for Surface Inspection Using Unmanned Aerial Vehicles

Wu,

Funabora,

Doki

et al. 2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

We implemented and evaluated our previous path planning method for inspection using unmanned aerial vehicles (UAVs) in real-world, and identified its shortcomings in handling positioning errors. Then, we proposed an enhanced method to address this problem. The previous method theoretically guaranteed complete coverage of targets and data quality. However, we verified it in bridge inspection experiments and found that the former has not been ensured. The crucial factors of data omission are clarified as the errors in UAV positioning. Our previous method relies on appropriately setting ideal allowances to counteract positioning errors, which is impractical. Therefore, we proposed an enhanced path planning method, which adaptively adjusts allowances according to positioning error to prevent omission while minimizing waypoints. In the simulation including positioning disturbances, the enhanced method consistently achieved full coverage in 1000 times simulation with over 28% waypoints less than the previous one.

show abstract

Section: Related Workmentioning

confidence: 99%