Logical methods of object recognition on satellite images using spatial constraints

TL;DR

This paper introduces a logical inference-based method for object recognition in satellite images, utilizing rules that describe object shapes to improve recognition accuracy.

Contribution

It presents a novel logical approach that applies rule-based inference for object recognition in satellite imagery.

Findings

Effective recognition of objects using logical inference.

Improved accuracy over traditional methods.

Applicable to various satellite image analysis tasks.

Abstract

A logical approach to object recognition on image is proposed. The main idea of the approach is to perform the object recognition as a logical inference on a set of rules describing an object shape.