Targeted change detection in remote sensing images
share
Recent developments in the remote sensing systems and image processing made it possible to propose a new method of the object classification and detection of the specific changes in the series of satellite Earth images (so called targeted change detection). In this paper we propose a formal problem statement that allows to use effectively the deep learning approach to analyze time-dependent series of remote sensing images. We also introduce a new framework for the development of deep learning models for targeted change detection and demonstrate some cases of business applications it can be used for.
READ FULL TEXT