ArcGIS Pro: Image Segmentation, Classification and Machine Learning

Extracting information from remotely sensed imagery is an important step to providing timely information for your GIS. ArcGIS Pro desktop provides a rich environment to process and exploit imagery. A significant aspect of this is the image segmentation and classification tools that process multispectral imagery, resulting in extracted feature data for the GIS. These can be combined with a range of machine learning algorithms for the automated feature extraction and identification, and also used as models that can be scaled up using Raster Analytics running on ArcGIS Image Server

[@reubengeorgeofoke2149]

Awesome, there is clear hope in Deep learning

[@vovin8132]

The final point is interesting; would projections alter object classification in deep learning analysis? Maybe a different projection might shift an object, like a ship or house, a couple pixels into a different shape.

[@nikibowe]

can i go this on a single band grey scale? also you didn't show us how to make a classification schema with out own datasets..

[@DrewPilant]

Very helpful- thank you! What fun!

[@SoItGoes1985]

At what part of the Classification Wizard do you add the additional rasters like a DEM of NDVI raster?

[@gistech8777]

good question

[@benmissimer]

Image processed is a worldview 3 scene. Where do I get one? The resolution on Landsat data is ROUGH! Is worldview the best / only source for data like this to play with?

[@SoItGoes1985]

Try NAIP 4-band from USGS EarthExplorer.

[@sadiksylla99]

i'm very happy to taht, thank you

[@steliosyiakoupis9869]

how to get a table for the erros and accuracy of our method?

[@pablogallegos6015]

I'm learning machine learning, I hope can solve problems in civil engineering.

[@WaqarAr]

Thank you.

[@alqu2222]

Though the subject is very interesting and popular, the presenter had too many pauses and Ahhhhs that caused the viewer to lose focus and it seems that the presenter was expecting something else than what he had on the screen which shows the ill preparation and sketchy approach of the video.