Peter Reinartz

TL;DR: In this paper, a generic ortho image processor is proposed to produce ortho images from airborne and spaceborne digital line scanner images, as well as images from frame cameras based on the Direct Georeferencing model using measurements of the exterior orientation of the sensor platform or sensor itself, the interior orientation (sensor parameters) and a digital elevation model.

TL;DR: In this article, a Symmetric Fully Convolutional Neural Network enhanced by Wavelet Transform is proposed to automatically carry out lane marking segmentation in aerial imagery, which can capture a large area in a short period of time by introducing an aerial lane marking dataset.

TL;DR: A deep learning architecture is proposed to classify hyperspectral remote sensing imagery by joint utilization of spectral–spatial information and the obtained results indicate the superiority of the proposed spectral-spatial deepLearning architecture against the conventional classification methods.

TL;DR: A new algorithm for automatic digital terrain model (DTM) generation from high resolution CARTOSAT-1 satellite images consists of two major steps: generation of digital surface models (DSM) from stereo scenes and hierarchical image filtering for DTM generation.

TL;DR: A novel disaster building damage monitoring method that combines the multispectral imagery and DSMs from stereo matching to obtain three kinds of changes and is applied to building change detection after the Haiti earthquake.