Bundle block adjustment is a fundamental concept in photogrammetry, involving the calculation of camera interior and exterior orientation parameters as well as the 3D coordinates of objects. Accurate block adjustment relies on Ground Control Points (GCPs), which are known points on the ground with precise coordinates. However, these concepts can be technical and often overlooked by users, leading to potential issues in drone or aerial photos data processing. This article demonstrates the importance of block adjustment and GCPs using ArcGIS PRO. To ensure clarity, this demonstration focuses on result exploration and avoids excessive mathematical explanations. For a more scientific understanding, please refer to my article on the fundamentals of photogrammetry. For scientific explanation, please refer to my article on the fundamentals of photogrammetry.
I use subset of aerial photos acquired with a Phase One iXS100F camera in Boise. The focal length of the camera is 50mm and the pixel size is 0.0046mm. Each photo has dimensions of 11608 rows by 8708 columns. The camera used in this project is not pre-calibrated, meaning the precise interior orientation parameters such as focal length and principal points need to be computed. Additionally, location and orientation of the camera needs to be known we need to be know to calculate 3D locations of objects in photogrammetry. The information generally called exterior orientation parameters could be stored in the exif file of the photo or provided separately as it is in this case. The rotation values are set to zero (omega, phi, kappa) for the purpose of this demonstration. The camera parameters and exterior orientation information are illustrated in the figure below.
With the interior and exterior orientation parameters determined, the next step is to create and configure an appropriate reality mapping workspace for the project. Detail documentation on orthomapping or reality mapping workspace in ArcGIS PRO can be found in this documentation.
Using the location information from the orientation file, the photos are approximately located in the appropriate space. However, without the rotation values, the photos are not correctly oriented. Some of the images in the collection shown below are oriented in the opposite direction.
The subsequent task is to perform block adjustment, which involves detecting key points in the images and calculating tie points, which are common points in areas of image overlap. Using triangulation techniques, the camera positions and orientations are computed from these tie points.
After bundle block adjustment, the images are relatively oriented correctly but still slightly misaligned with the reference basemap, as shown in the illustration below.