# Patched Line Segment Learning for Vector Road Mapping

arXiv: https://arxiv.org/abs/2309.02923 · HTML: https://mlsanity.com/paper/2309.02923/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 37 references that are in the map.
- First posted on arXiv 6 Sep 2023.

## Builds on (most cited first, 16 of 16)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://mlsanity.com/paper/1606.04797/) (2016)
- [SpaceNet: A Remote Sensing Dataset and Challenge Series](https://mlsanity.com/paper/1807.01232/) (2018)
- [RoadTracer: Automatic Extraction of Road Networks from Aerial Images](https://mlsanity.com/paper/1802.03680/) (2018)
- [Relationformer: A Unified Framework for Image-to-Graph Generation](https://mlsanity.com/paper/2203.10202/) (2022)
- [Holistically-Attracted Wireframe Parsing](https://mlsanity.com/paper/2003.01663/) (2020)
- [Sat2Graph: Road Graph Extraction through Graph-Tensor Encoding](https://mlsanity.com/paper/2007.09547/) (2020)
- [Learning Attraction Field Representation for Robust Line Segment Detection](https://mlsanity.com/paper/1812.02122/) (2018)
- [PolyWorld: Polygonal Building Extraction with Graph Neural Networks in Satellite Images](https://mlsanity.com/paper/2111.15491/) (2021)
- [RNGDet: Road Network Graph Detection by Transformer in Aerial Images](https://mlsanity.com/paper/2202.07824/) (2022)
- [Holistically-Attracted Wireframe Parsing: From Supervised to Self-Supervised Learning](https://mlsanity.com/paper/2210.12971/) (2022)
- [RNGDet++: Road Network Graph Detection by Transformer with Instance Segmentation and Multi-scale Features Enhancement](https://mlsanity.com/paper/2209.10150/) (2022)
- [iCurb: Imitation Learning-based Detection of Road Curbs using Aerial Images for Autonomous Driving](https://mlsanity.com/paper/2103.17118/) (2021)
- [Learning and Aggregating Lane Graphs for Urban Automated Driving](https://mlsanity.com/paper/2302.06175/) (2023)
- [Learning Regional Attraction for Line Segment Detection](https://mlsanity.com/paper/1912.09344/) (2019)
- [PaRK-Detect: Towards Efficient Multi-Task Satellite Imagery Road Extraction via Patch-Wise Keypoints Detection](https://mlsanity.com/paper/2302.13263/) (2023)

## Built on by (most cited first, 5 of 5)

- [Towards Satellite Image Road Graph Extraction: A Global-Scale Dataset and A Novel Method](https://mlsanity.com/paper/2411.16733/) (2024)
- [ScaleLSD: Scalable Deep Line Segment Detection Streamlined](https://mlsanity.com/paper/2506.09369/) (2025)
- [UniVector: Unified Vector Extraction via Instance-Geometry Interaction](https://mlsanity.com/paper/2510.13234/) (2025)
- [DOGE: Differentiable Bezier Graph Optimization for Road Network Extraction](https://mlsanity.com/paper/2511.19850/) (2025)
- [Beyond Endpoints: Path-Centric Reasoning for Vectorized Off-Road Network Extraction](https://mlsanity.com/paper/2512.10416/) (2025)

## Abstract (arXiv)

This paper presents a novel approach to computing vector road maps from satellite remotely sensed images, building upon a well-defined Patched Line Segment (PaLiS) representation for road graphs that holds geometric significance. Unlike prevailing methods that derive road vector representations from satellite images using binary masks or keypoints, our method employs line segments. These segments not only convey road locations but also capture their orientations, making them a robust choice for representation. More precisely, given an input image, we divide it into non-overlapping patches and predict a suitable line segment within each patch. This strategy enables us to capture spatial and structural cues from these patch-based line segments, simplifying the process of constructing the road network graph without the necessity of additional neural networks for connectivity. In our experiments, we demonstrate how an effective representation of a road graph significantly enhances the performance of vector road mapping on established benchmarks, without requiring extensive modifications to the neural network architecture. Furthermore, our method achieves state-of-the-art performance with just 6 GPU hours of training, leading to a substantial 32-fold reduction in training costs in terms of GPU hours.