ABSTRACT
This chapter introduces how to produce one synthetic remote sensing image from multiple sources, and in making use directly of TensorFlow to train the model from Python code. It describes a method which performs the gapfilling of a cloudy optical image, from multiple optical images and a synthetic aperture radar image. The dataset used during this exercise consisted of three optical Sentinel-2 images, including one scene polluted by clouds, and one synthetic aperture radar Sentinel-1 image. It should be noted that the images acquired respectively before and after the cloudy images are not completely free of clouds: one could extend this exercise to the retrieval of the missing contents in each of these images, repeating the steps described in this section with each image as a target estimated image, one by one.
