Torsten Büschenfeld

Learn More
In this article we present a method that extracts plantations from satellite imagery by finding and exploiting appropriate feature space projections. Segmentation is done using an automatic two-region segmentation based on the level set method. The behaviour of this algorithm is defined by a statistical region model that describes the similarity of regions(More)
— ViPiD is a complete framework for audio and 3D video capturing of one or several moving persons as well as the creation of 3D person models for intuitive dialog systems. Therefore we are setting up a multi-camera environment for 3D scene analysis, incorporating aspects such as 3D/4D reconstruction , motion estimation, virtual camera integration, coding of(More)
Recently, a lot of classification methods within GIS applications rely on object-based image analysis. The image is partitioned by a low-level segmentation method, and classification is processed on each segment individually. Often, the resulting homogeneous segments do not allow for a correct classification. Hence, a semantic processing considering(More)
In this paper, we present a method for automatic refinement of training data. Many classifiers from machine learning used in applications in the remote sensing domain, rely on previously labelled training data. This labelling is often done by human operators and is bound to time constraints. Hence, selection of training data must be kept practical which(More)
  • 1