# Targeted change detection in remote sensing images

## TL;DR

This paper introduces a formal problem statement and a new deep learning framework for targeted change detection in remote sensing images, enabling effective analysis of time series for specific object changes with practical business applications.

## Contribution

It presents a formal problem formulation and a novel deep learning framework tailored for targeted change detection in satellite imagery.

## Key findings

- Demonstrated the framework on real-world cases
- Showed effectiveness in detecting specific changes over time
- Highlighted potential business applications

## Abstract

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.