Multimodal urban remote sensing image registration via roadcross triangular feature

HIGHLIGHTS

• who: -Image, matching and Multimodal Urban, Remote Sensing from the Vegas, Paris, Shanghai, and Khartoum datasets to train our model of road network extractionIn order to improve the ability of generalization, the selected datasets and road labels contain a large number of road remote sensing and aerial imagery at different times of the world, and cover a huge geographic area from , to , square kilometer. Training images with three channels and labels are tiled into , u00d7 , pixel chips and the image resolution is about , cm/pixel. Test image pairs: The experiment test image pairs include three kinds of . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.