Premium
Semi‐Automated Registration Of Close‐Range Hyperspectral Scans Using Oriented Digital Camera Imagery And A 3d Model
Author(s) -
Sima Aleksandra A.,
Buckley Simon J.,
Kurz Tobias H.,
Schneider Danilo
Publication year - 2014
Publication title -
the photogrammetric record
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.638
H-Index - 51
eISSN - 1477-9730
pISSN - 0031-868X
DOI - 10.1111/phor.12049
Subject(s) - artificial intelligence , computer vision , hyperspectral imaging , computer science , scale invariant feature transform , bundle adjustment , orientation (vector space) , remote sensing , feature extraction , mathematics , photogrammetry , geography , geometry
Diverse applications can benefit from the integration of data acquired by a new generation of close‐range imaging sensors with high‐resolution three‐dimensional (3D) geometric data. However, such integration requires increased automation and efficiency of image‐data registration to guarantee adoption by users beyond the geomatics community. This paper presents a semi‐automated method for registering terrestrial panoramic hyperspectral imagery with lidar models and conventional digital photography. The method relies on finding corresponding points between images acquired in significantly different parts of the electromagnetic spectrum, from different viewpoints, and with different spatial resolution and geometric projections. Optimisation of the scale invariant feature transform ( SIFT ) operator was required to ensure a sufficient number of homologous points, as well as a routine for eliminating false matches. A band selection routine maximises the number of points found while minimising the input data for SIFT . Three‐dimensional object coordinates were derived in the lidar model and used as control points in a bundle block adjustment to determine the hyperspectral exterior orientation and intrinsic camera parameters. The method developed was applied to two datasets with different characteristics, and the results indicate that the proposed method is a time‐saving alternative to manual approaches.