Skip to content

A simple conv-deconv neural network for semantic segmentation on satellite data

Notifications You must be signed in to change notification settings

ptalon91/cnn-for-satellite-data

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 

Repository files navigation

cnn-for-satellite-data

A simple conv-deconv neural network for semantic segmentation on satellite data

The objective of this project is to build a simple convolution-deconvolution neural network model with Keras (TensorFlow engine backend), that can process satellite data and generate a semantic segmentation.

The data that will be used to elaborate and calibrate this model is the Vaihingen dataset. Description and download available at http://www2.isprs.org/commissions/comm3/wg4/detection-and-reconstruction.html

Prerequisites:

Steps:

  • Data preprocessing (crop series of smaller images from the patches, determine the preponderant label for each image, for training and validation data).
  • Simple convolutional neural network (Make the model predict a single label per image).
  • Simple convolution-deconvolution neural network (Make the model predict a label for each pixel of the image).
  • Evaluate the performances.

About

A simple conv-deconv neural network for semantic segmentation on satellite data

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages