Question

Machine learning ( roofs)

  • 15 March 2020
  • 3 replies
  • 17 views

Badge

Hello everybody,

I found out that FME includes machine learning in the 2019 version so I start exploring and I found a few great examples.

I read this article and it was great.

I have one task and if anybody has some ideas, some tips, some suggestions, I would be grateful.

The task is: I have 2 satellite images (panchromatic has 0,46 m spatial resolution ) and ( multispectral image has 1,84 m spatial resolution ). I want with machine learning methods classified roofs on the image ( i want to figure out roof orientation, roof area and roof slope ).


3 replies

Badge +2

Hi @maloks,

Great question! @dmitribagh just talked about using machine learning to detect roofs in the FME 2020 Unleashed: Authoring webinar so that might be a good place to get some ideas. He discusses this topic in the Machine learning and cognitive services section - the roof demo using the PicterraConnector starts around the 51 minute mark.

Badge

Hi @maloks,

Great question! @dmitribagh just talked about using machine learning to detect roofs in the FME 2020 Unleashed: Authoring webinar so that might be a good place to get some ideas. He discusses this topic in the Machine learning and cognitive services section - the roof demo using the PicterraConnector starts around the 51 minute mark.

I'm fascinated by the possibilities offered by FME, Picterra. I can't wait to try it out. Thanks a lot for your answer and suggestion, @chrisatsafe .

Badge +2

I'm fascinated by the possibilities offered by FME, Picterra. I can't wait to try it out. Thanks a lot for your answer and suggestion, @chrisatsafe .

That's great, we'd love to hear how it works out for you! Feel free to update this thread after you have had the chance to give it a try :)

Reply