کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
246422 | 502368 | 2015 | 20 صفحه PDF | دانلود رایگان |
• Three infrastructureless tracking systems are developed and evaluated under different scenarios.
• The first system is an inertial navigation system using an IMU.
• The second system uses a monocular camera and SfM algorithms.
• The third system fuses inertial and vision using an extended Kalman filter.
• The joint inertial–vision tracking system is proven promising in locating users on construction sites.
Current research trends in the field of automation in construction focus on automating traditional tasks with the intent of improving the accuracy and streamlining the flow of information throughout the construction process. One viable, potentially automated task is site inspection. This consists of a set of vital yet notoriously laborious, costly, time-consuming, and error-prone activities, when performed manually by site inspectors. Incorporating recent contributions from the fields of inertial-based step detection and vision-based egomotion estimation, this paper develops and evaluates infrastructureless inertial and visual systems to ubiquitously localize people in construction environments. The implementation and the results obtained from validation experiments demonstrate that a lightweight, portable, and cost-effective hybrid inertial–vision tracking system provides the most promise for accurate, precise, and robust localization of mobile users in construction environments. It achieves, on average, an accuracy of around 1% of the total traveled distance, compared to the INS and vision systems that provide an accuracy of less than 2% and 3% of the total traveled distance respectively.
Journal: Automation in Construction - Volume 56, August 2015, Pages 47–66