Visual radar combined unmanned aerial vehicle positioning method based on multi-modal scale normalization
The invention discloses a visual radar and unmanned aerial vehicle combined positioning method based on multi-modal scale normalization, and the method comprises the steps: collecting visual radar multi-modal data, including an unmanned aerial vehicle image and target distance information provided b...
Gespeichert in:
Hauptverfasser: | , , , , , |
---|---|
Format: | Patent |
Sprache: | chi ; eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The invention discloses a visual radar and unmanned aerial vehicle combined positioning method based on multi-modal scale normalization, and the method comprises the steps: collecting visual radar multi-modal data, including an unmanned aerial vehicle image and target distance information provided by a radar; marking corresponding bounding boxes according to different positions of the unmanned aerial vehicle in the image; recombining the unmanned aerial vehicle image slices into a batch two-dimensional image slice structure according to the target distance information by adopting a distance sensing image adaptive slice network to obtain training data; constructing a scale normalization network based on modal fusion, and training the normalization network by using the training data; constructing an unmanned aerial vehicle positioning visual model, and training the visual model by using the training data; and inputting a test unmanned aerial vehicle image and target distance information into the distance sensin |
---|