Development Of The Ai ​​Procedural Adaptation By Implementing A Method For The (Pre) Training Of A Convolutionary Neural Network (U-Net, Neural Networks) In The Tool For The Derivation Of Tree Locations

  Address - Germany

  Contact details - 565656565

  Tender notice no. - 76454545

  GT Ref Id - 108244222

  Document Type - Tender Notices

Description - Description: In the past, the regional association Ruhr has several open source tools to derive buildings, green roofs and tree locations from aerial photo data. These were made available as a download after the public (https://github.com/mundialis/r vr_interface).When deriving tree locations, the classification and regression violation of Random Forest is currently used for machine learning as a sub-area of ​​artificial intelligence (AI). The model can be easily applied to homogeneous sub -rooms, but is subject to larger areas of methodological limitations due to the heterogeneity of the input data (vegetation, image flight year, sensors). To expand the AI ​​approach, the methodology

Gt Ref Id - 108244222

Deadline - May 05, 2026