Improving HySpex Sensor Co-registration Accuracy using BRISK and Sensor-model based RANSAC

Abstract

ABSTRACT:In this paper a method to improve the co-registration accuracy of two separate HySpex SWIR and VNIR cameras is proposed. The first step of the presented approach deals with the detection of point features from both scenes using the BRISK feature detector. After matching these features, the match coordinates in the VNIR scene are orthorectified and the resulting ground control points in the SWIR scene are filtered using a sensor-model based RANSAC. This implementation of RANSAC estimates the boresight angles of a scene by iteratively fitting the sensor-model to a subset of the matches. The boresight angles which can be applied to most of the remaining matches are then used to orthorectify the scene. Compared to previously used methods, the main advantages of this approach are the high robustness against outliers and the reduced runtime. The proposed methodology was evaluated using a test data set and it is shown in this work that the use of BRISK for feature detection followed by sensor-model based RANSAC significantly improves the co-registration accuracy of the imagery produced by the two HySpex sensors.

Hyperspectral imaging of art: Acquisition and calibration workflows

“…However, even after geometric calibration, small errors can still exist due to minor optical aberrations etc. Co-registration can be optimized by several methods, such as, for example, feature-based matching over an overlapping spectral range (Schwind et al 2014), which can provide an automated and robust result with sub-pixel level accuracy.…”

Differences and Similarities in the Processing of Airborne and Spaceborne Hyperspectral Data Shown on HySpex and EnMap Processing Chains

“…Then tie points between VNIR in SWIR data are found in bands located in the overlapping spectral domain using a BRISK matching. [10] After this step, the VNIR image is orthorectified using the physical sensor model, the GPS-/IMU-data, the previously obtained mounting angles and the DEM by the DLR software ORTHO. The tie points are inserted as check points (CPs) to obtain the respective geo-coordinates.…”

Evaluating the applicability of BRISK for the geometric registration of remote sensing images

“…HySpex (Schwind, Schneider, and Müller 2014)) and even modern satellite platforms (e.g. Using satellite orientation and position measurements alone is often not sufficient to achieve the high geometric accuracies required for many applications such as image fusion, change detection or digital elevation model (DEM) generation.…”

“…Consequently, we already used BRISK, followed by a sensormodel-based RANSAC (RANdom SAmple Consensus, (Fischler and Bolles 1981)) to improve the co-registration accuracy of two HySpex cameras (Schwind, Schneider, and Müller 2014). Consequently, we already used BRISK, followed by a sensormodel-based RANSAC (RANdom SAmple Consensus, (Fischler and Bolles 1981)) to improve the co-registration accuracy of two HySpex cameras (Schwind, Schneider, and Müller 2014).…”