Optimizing Bundle Block Adjustment for High-Overlap Small-Format Multi-Head Camera Systems

Abstract. Small-format multi-head camera systems integrating nadir and oblique cameras are becoming standard in photogrammetry, offering the versatility needed for UAV-based mapping and surveying applications to collect high-quality data in challenging environments. A key consideration for these systems is the inherent limitations due to their small size, which may lead to various trade-offs, such as reduced ground coverage, overlap, and image quality, as well as constraints on using high-precision instruments like GNSS and INS. In this contribution, we propose the compass direction (CD) frame as a superior reference frame to avoid confusion regarding oblique cameras and oblique viewing directions. This approach supports continuous and consistent ground coverage, similar to the nadir camera, and facilitates easier analysis. We focus particularly on relative orientation, as it is a decisive factor for the consistency of subsequently derived information. We evaluate various influencing factors, including image connectivity, image ray intersection angle, and tie point multiplicity, in both single-camera (nadir) and penta-camera (nadir and oblique). The analysis reveals significant deviations in image connectivity from proportionality to image overlap. The percentage of overlap and side lap for nadir images influences the overlap for oblique cameras. While, there are no gaps at the near line between overlapping frames in each individual compass direction, the nadir images overlaps do not result in the same symmetry in the oblique image overlaps. Achieving such symmetry could enhance the visibility of facades in oblique images. In terms of the image ray intersection angle, the maximum intersection angles per tie point are well below the theoretical maximum, indicating a moderate exploitation of the accuracy potential offered by oblique images. Furthermore, analysis of the RMS error suggests that by adding the oblique images to the nadir image block and removing the 2-fold tie points, vertical accuracy improved considerably, while planimetric accuracy remained consistent.

Authors

Institutions

Publication Details

Journal
˜The œinternational archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences
Published
2024-12-13
DOI
https://doi.org/10.5194/isprs-archives-xlviii-2-w7-2024-17-2024
Primary Topic
3D Surveying and Cultural Heritage
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Optimizing Bundle Block Adjustment for High-Overlap Small-Format Multi-Head Camera Systems

Wilfried Karel, Camillo Ressl, Norbert Pfeifer, Thirawat Bannakulpiphat
˜The œinternational archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences
3D Surveying and Cultural Heritage
article

Optimizing Bundle Block Adjustment for High-Overlap Small-Format Multi-Head Camera Systems

Wilfried Karel, Camillo Ressl, Norbert Pfeifer, Thirawat Bannakulpiphat
article en

Abstract

Abstract. Small-format multi-head camera systems integrating nadir and oblique cameras are becoming standard in photogrammetry, offering the versatility needed for UAV-based mapping and surveying applications to collect high-quality data in challenging environments. A key consideration for these systems is the inherent limitations due to their small size, which may lead to various trade-offs, such as reduced ground coverage, overlap, and image quality, as well as constraints on using high-precision instruments like GNSS and INS. In this contribution, we propose the compass direction (CD) frame as a superior reference frame to avoid confusion regarding oblique cameras and oblique viewing directions. This approach supports continuous and consistent ground coverage, similar to the nadir camera, and facilitates easier analysis. We focus particularly on relative orientation, as it is a decisive factor for the consistency of subsequently derived information. We evaluate various influencing factors, including image connectivity, image ray intersection angle, and tie point multiplicity, in both single-camera (nadir) and penta-camera (nadir and oblique). The analysis reveals significant deviations in image connectivity from proportionality to image overlap. The percentage of overlap and side lap for nadir images influences the overlap for oblique cameras. While, there are no gaps at the near line between overlapping frames in each individual compass direction, the nadir images overlaps do not result in the same symmetry in the oblique image overlaps. Achieving such symmetry could enhance the visibility of facades in oblique images. In terms of the image ray intersection angle, the maximum intersection angles per tie point are well below the theoretical maximum, indicating a moderate exploitation of the accuracy potential offered by oblique images. Furthermore, analysis of the RMS error suggests that by adding the oblique images to the nadir image block and removing the 2-fold tie points, vertical accuracy improved considerably, while planimetric accuracy remained consistent.

˜The œinternational archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciencesVol. XLVIII-2/W7-2024
TU Wien (AT), Chulalongkorn University (TH)
Chulalongkorn University
Sustainable cities and communities
Openalex Percentile: Top 53%
3D Surveying and Cultural Heritage
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.