Jarron, DavidLichti, Derek D.Shahbazi, Mozhdeh M.Radovanovic, Robert S.2019-07-042019-07-042019-01Jarron, D., Lichti, D. D., Shahbazi, M. M., & Radovanovic, R. S. (2019). Multi-camera panormamic imaging system calibration. In Proceedings of the 11th International Conference on Mobile Mapping. Shenzhen, China, 6-8 May 2019. Shenzhen: International Conference on Mobile Mapping.http://hdl.handle.net/1880/110580https://dx.doi.org/10.11575/PRISM/36700A mobile mapping system (MMS) is a three-dimensional reality capture system that collects georeferenced spatial data with integrated navigation and imaging sensors from a moving vehicle. Several imaging subsystems can be found on board an MMS, such as panoramic camera systems and LiDAR sensors. The data collected from a panoramic imaging system must be accurately georeferenced and the sensors must be rigorously calibrated to ensure accurate registration of images to the point clouds collected by the LiDAR sensors, and to ensure panoramic images are generated seamlessly. The panoramic imaging system studied in this work is the Ladybug5 (by FLIR Integrated Imaging Solutions), which is a spherical camera system comprised of six individual wide-angle cameras. Having accurate estimates of the interior and relative orientation parameters of these cameras is essential for integrating the camera system with other sensors in the MMS to generate georeferenced spatial data. However, field experience has shown that factory-provided calibrations may be insufficiently accurate for high-precision applications. An investigation of the geometric calibration of the Ladybug 5 system was conducted in a dedicated indoor calibration facility at the University of Calgary: an 11 m x 11 m x 4 m field comprising 291 signalized photogrammetry targets. Multiple free-network, self-calibrating bundle adjustments were performed using different sets of constraints to model several systematic error sources. Weighted constraints were included in the adjustment to enforce the stability of the six relative orientation parameters between image pairs, and separate colour channel adjustments were used to compensate for chromatic aberrations. The overall fit of observations to the calibration model as measured by the root mean square error of the image point residuals was at the level of 0.3-0.4 pixels. Mean object point precision was at the 0.3 mm level. Rectified and ortho-rectified panoramas were also generated to verify the calibrations precision and observe how adjustments with constraints effect panorama generation.engUnless otherwise indicated, this material is protected by copyright and has been made available with authorization from the copyright owner. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.Ladybug CameraMobile MappingCamera CalibrationPanoramic Imagingconference paperCRDPJ505367-16