DVZ Datenverarbeitungszentrum Mecklenburg-Vorpommern
19059 Schwerin, Germany
High-resolution geodata with different usage aspects are used in numerous applications and specialist procedures that we develop for the state of Mecklenburg-Western Pomerania .
The task
For example, neural networks can compare buildings and their changes using time-delayed geodata and also detect deviations from development specifications. Deviations between development specifications and the actual development can be identified using AI techniques such as machine learning and deep learning.
our range
Your task is to create a suitable neural network that can be used to identify buildings and their changes on parcels of land. The training of such a neural network is based on high-resolution aerial photographs, which are then prepared as a data set, which can be done with GPU and TPU support.
The following tasks may vary depending on scope, time and focus:
analysis and processing of requirements
research for various solutions
Creation of a dataset based...