ICDAR 2021 Competition on Historical Map Segmentation

TL;DR

This paper reports on the ICDAR 2021 Competition on Historical Map Segmentation, showcasing advances in detecting map features using deep learning and image processing techniques across three distinct tasks.

Contribution

It introduces a standardized dataset and evaluation framework for historical map segmentation, and demonstrates state-of-the-art methods for building detection, content segmentation, and intersection point localization.

Findings

Task 1 achieved high accuracy with DenseNet-121 in weakly supervised training.

Task 2 improved segmentation edge detection with a U-Net-like FCN and binarization.

Task 3 effectively located intersection points using a combined pipeline with line detection and template matching.

Abstract

This paper presents the final results of the ICDAR 2021 Competition on Historical Map Segmentation (MapSeg), encouraging research on a series of historical atlases of Paris, France, drawn at 1/5000 scale between 1894 and 1937. The competition featured three tasks, awarded separately. Task~1 consists in detecting building blocks and was won by the L3IRIS team using a DenseNet-121 network trained in a weakly supervised fashion. This task is evaluated on 3 large images containing hundreds of shapes to detect. Task~2 consists in segmenting map content from the larger map sheet, and was won by the UWB team using a U-Net-like FCN combined with a binarization method to increase detection edge accuracy. Task~3 consists in locating intersection points of geo-referencing lines, and was also won by the UWB team who used a dedicated pipeline combining binarization, line detection with Hough transform, candidate filtering, and template matching for intersection refinement. Tasks~2 and~3 are evaluated on 95 map sheets with complex content. Dataset, evaluation tools and results are available under permissive licensing at \url{https://icdar21-mapseg.github.io/}.