Towards automating the nautical chart generalization workflow

doi:10.1080/10095020.2024.2366873

Towards automating the nautical chart generalization workflow

2024-06-19

1

PRE

AI

DOI：10.1080/10095020.2024.2366873

Original

Original request for help

Save

Abstract

En 中文

Current nautical chart generalization methods are notably labor intensive, requiring significant levels of human intervention to compile, update, and maintain chart products. The ideal situation would be a fully automated solution for generating nautical charts seamlessly from a comprehensive database, on demand, at the appropriate scale, at the point of use, and respecting the product constraints. However, regardless of the various research efforts and advancements in technology, including those involving AI, nautical chart generalization tasks are still performed manually, or semi-manually, where a likelihood of human error is expected. This manuscript presents a research effort toward automated chart compilation through scales. Nautical chart generalization guidelines are extracted, categorized, and translated into machine readable rules, utilized by a multi-agent model to perform the generalization of the source data to the target scale with no topological violations. This is illustrated in three testbeds for the most important ENC feature classes. While topology is maintained, the model utilizes readily available algorithms that, generally, compromise safety. Therefore, a custom validation tool detects safety violations for user intervention. The model has been made flexible to incorporate algorithms that align with application constraints, especially safety, as they become available.

Keywords:

ENC compilation

automated generalization

nautical chart generalization

ocean mapping

safety of navigation

multi-agent modeling

Towards automating the nautical chart generalization workflow

Abstract

Journal

Organization