# Robot Localization in Floor Plans Using a Room Layout Edge Extraction Network

## TL;DR

This paper presents a monocular camera-based indoor robot localization method that uses a neural network to extract room layout edges from images and a particle filter to match these edges to a floor plan, enabling efficient and accurate pose estimation.

## Contribution

The authors introduce a novel, efficient localization system combining CNN-based edge extraction and particle filtering for floor plan matching, reducing reliance on specialized sensors and manual data collection.

## Key findings

- Achieves robust and accurate indoor localization in real-world experiments.
- Operates efficiently with low computational and memory requirements.
- Does not require sensor data beyond a monocular camera.

## Abstract

Indoor localization is one of the crucial enablers for deployment of service robots. Although several successful techniques for indoor localization have been proposed, the majority of them relies on maps generated from data gathered with the same sensor modality used for localization. Typically, tedious labor by experts is needed to acquire this data, thus limiting the readiness of the system as well as its ease of installation for inexperienced operators. In this paper, we propose a memory and computationally efficient monocular camera-based localization system that allows a robot to estimate its pose given an architectural floor plan. Our method employs a convolutional neural network to predict room layout edges from a single camera image and estimates the robot pose using a particle filter that matches the extracted edges to the given floor plan. We evaluate our localization system using multiple real-world experiments and demonstrate that it has the robustness and accuracy required for reliable indoor navigation.