CoChat
Sign Up
Cited by 303D Modeling in Geospatial ApplicationsUrban Design and Spatial AnalysisGeographic Information Systems Studies

The DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index; none of the 7 sources its claims cite is checked for support yet.

An indoor navigation model and its network extraction

Open access
Mortari, Filippo, Clementini, Eliseo, Zlatanova, Sisi, Liu, Liu·2019·Applied Geomatics
DOI 10.1007/s12518-019-00273-8
Open this paper in CoChat — free accountOpen PDF

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

The paper proposes an extension of the IndoorGML navigation module that imports CityGML LOD4 semantics and an improved geometric network model (IGNM) for automatic indoor navigation network extraction. IGNM combines medial-axis and visibility-graph ideas via a two-level hierarchical algorithm and is evaluated against Door-to-Door on indoor floor plans.

Key findings

  • The extended IndoorGML navigation module refines concepts such as `ConnectedSpace` into `Window` and `Door`, adds classes including `Building`, `Storey`, `NavigableSpace`, and `NonNavigableSpace`, and is designed to import CityGML LOD4 data without information loss while adding semantics for human navigation.
  • IGNM preserves the duality between primal and dual space, incorporates semantic information, and uses a two-level strategy: a coarse inter-space weighted graph with Dijkstra, then a local intra-space path built through inward offsetting, boundary sampling, constrained Delaunay triangulation, crossing-facet subdivision, node displacement, and node linking.
  • In the reported comparison, IGNM produced 77 nodes and 152 edges versus Door-to-Door’s 24 nodes and 194 edges; the best path length was 46.21 m for IGNM versus 25.57 m for Door-to-Door, and graph construction plus path finding took 0.002 versus 0.005. The text states IGNM’s path was ≈1.8 times longer in that case and ≈1.6 times longer on average, with ≈3.2 more nodes and ≈0.78 times the edges.
  • IGNM is presented as avoiding S-MAT distortions in open spaces and Door-to-Door’s high edge count and complex route descriptions, while improving localization through finer resolution and bidirectional primal–dual mapping.

Methods

  • Design: an extension of IndoorGML’s Indoor Navigation Module with CityGML LOD4 concepts, plus the IGNM automatic network extraction algorithm as a hybrid of S-MAT and Door-to-Door approaches.
  • Data: the OTB Building (TU Delft) floor plan drawn with Java OpenStreetMap Editor (JOSM) and the FJK-House CityGML LOD4 dataset provided by Karlsruhe Institute of Technology; for FJK-House, 2D footprints of each floor were extracted before running the algorithm.
  • Analysis: networks and shortest paths were constructed and compared with Door-to-Door using Dijkstra, reporting number of nodes, number of edges, best path length, and graph construction plus path-finding time (Table 1).

Limitations

  • Stated: further work is needed for peculiar indoor shapes that produce non-straight paths and for cases where inward offsetting produces non-convex polygons; optimization is needed to reduce the current number of nodes and edges; experiments on 3D networks are still needed; and interconnecting 2D floor networks into 3D navigation networks remains to be investigated. The paper also notes that corner-located doors can produce serpentine routes requiring path simplification, and that the centroid-outside-contour case for inset polygons has not yet been implemented.
  • Stated/evident: the evaluation uses only two datasets (OTB Building and FJK-House), and Table 1 compares only against Door-to-Door, so broader validation across building types and other network construction techniques is not shown.
  • Evident: the reported IGNM best path is longer than the optimal Door-to-Door path (46.21 m vs 25.57 m; ≈1.6–1.8× longer), indicating a trade-off in path optimality.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1007/s12518-019-00273-8

Cites

36 of 36 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 of 7 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    The IGNM-generated shortest path is approximately 1.8 times longer than the Door-to-Door path.

    “regarding the length of shortest path, IGNM generated path is ≈ 1.8 times longer than Door-to-Door which is known to be optimal.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Door-to-Door path finding is optimal.
  • Findingown data

    In the average case, the IGNM path is approximately 1.6 times longer than the optimal Door-to-Door path.

    “In average case, IGNM proves to be ≈ 1.6 longer than the optimal Door-to-Door, which is a promising result.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Door-to-Door path finding is optimal.
  • Findingown data

    The algorithm might produce serpentine routes in particular geometric environments such as doors located in a corner of a room, due to the shape of the triangulation.

    “There are particular geometric environments, such as doors located in a corner of a room, where the algorithm might produce some serpentine routes due to the shape of the triangulation (see Fig. 19a).”
    Quote in the paperNot a substantial claim
  • Findingown data

    The number of nodes produced by IGNM is approximately 3.2 times higher than the number produced by Door-to-Door.

    “the number of nodes produced with IGNM approach is higher than Door-to-Door (≈ 3.2 more)”
    Quote in the paperSubstantial claim
  • Findingown data

    The number of edges produced by IGNM is approximately 0.78 times the number produced by Door-to-Door.

    “the amount of edges is lower (≈ 0.78 times).”
    Quote in the paperSubstantial claim
  • It may state two separate claims as one. 14
  • Probably not a substantial claim. 16
  • The quote only partly states this claim. 1

  • It may state two separate claims as one.

    Findingown data

    The intra-space connectivity algorithm works correctly in arbitrarily shaped rooms with concave corners, rooms with obstacles, and rooms with open spaces that generate convex and concave inset polygons.

    “The intra-space connectivity algorithm proved to work correctly in arbitrarily shaped rooms with concave corners, rooms with obstacles, rooms with open spaces such that convex and concave inset polygons are generated.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Findingown data

    In the S-MAT of the hallway of OTB (TUDelft) building, detours are generated and paths are distorted towards the middle of the open space.

    “Figure 8 shows S-MAT of the hallway of OTB (TUDelft) building: detours are generated, and paths are distorted towards the middle of the open space.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Premisecitation

    S-MAT is a centerline-based topology network which can abstract linear features from simple polygons.

    “S-MAT is a centerline-based topology network which can abstract linear features from simple polygons.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lee 2004 source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisenone given

    S-MAT allows indoor route calculation to be conducted readily and fast.

    “The main benefits of S-MAT can be summarized as follows: indoor route calculation can be conducted readily and fast;”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    The extension introduced the semantics of various fixtures in indoor space of buildings, making the navigation model more suitable for human needs.

    “introduced the semantics of various fixtures in indoor space of buildings making the navigation model more suitable for human needs.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Premisecitation

    Medial axis approaches are a good representation of human navigation in narrow corridors, especially to avoid obstacles, but introduce distortions in open space.

    “Normally, medial axes approaches are a good representation of human navigation in narrow corridors, especially to avoid obstacles, but introduce distortions in open space.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Lee 2004 source

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Premisenone given

    Visibility approaches work better in open spaces.

    “On the other hand, visibility approaches work better in open spaces.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    The resulting network from the extraction technique takes advantages of both medial axis and visibility approaches and better mimics human beings' navigation in indoor environments.

    “In our extraction technique, the resulting network takes advantages of both approaches and better mimics human beings' navigation in indoor environments.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Interpretationnone given

    The graph is not only topological but also geometric: nodes are georeferenced in a 3D context, and edges contain metric information.

    “The graph is not only topological but also geometric: nodes have been georeferenced in a 3D context, while edges contain metric information.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.

    Interpretationnone given

    Since the intermediate points of a path between a source and a target are only the concave corners, the tracking of a subject or object is not efficient.

    “Since the intermediate points of a path between a source and a target are only the concave corners, the tracking of a subject or object is not efficient.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationnone given

    The geometric algorithm is local and capable of supporting dynamic changes.

    “the geometric algorithm is local and capable of supporting dynamic changes.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationnone given

    In IGNM, obstacles are encompassed and allowed to be inserted, removed, or displaced as required at run-time.

    “since obstacles not only are encompassed but are allowed to be inserted, removed, or displaced as required at run-time.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Premisenone given

    Two nodes are considered mutually visible if their direct connection does not intersect any of the region's boundaries.

    “Two nodes are considered as mutually visible if their direct connection does not intersect any of the region's boundaries (see Fig. 9).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisenone given

    In visibility-based networks, the amount of space required for storing nodes and edges is quadratic in convex spaces with no obstacles.

    “The amount of space required for storing nodes and edges: primarily in convex spaces with no obstacles, the number of possible combinations is quadratic (please refer to Fig. 10 to get a holistic idea).”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationown data

    S-MAT may create weird networks for corridors and fails for large arbitrarily shaped spaces.

    “Despite the above benefits, we could also identify several drawbacks, since it may create weird networks for corridors and it fails for large arbitrarily shaped spaces.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Premisenone given

    A centerline approach gives a good representation of natural human behavior within indoor environments.

    “a centerline approach gives a good representation of natural human behavior within indoor environments.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.

    Interpretationreasoning

    Door-to-door route descriptions may be very complex to be interpreted by humans.

    “Route descriptions may result very complex to be interpreted by humans.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Premisecitation

    In visibility-based networks, there are numerous possible networks based on the different routes a person can take, unlike centerline-based approaches where there is only one consistent network for the whole building.

    “As discussed in Taneja et al. (2011)), unlike the centerline-based approaches, where there is only one consistent network for the whole building, in visibility-based networks, there are numerous networks that are possible, based on the different routes a person can take.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Taneja et al. (2011) source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.

    Interpretationnone given

    The CDT can generate a centerline-based path similarly to S-MAT.

    “the CDT can generate a centerline-based path, similarly to S-MAT.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    A finer-grained partitioning of indoor navigable space can significantly contribute to facilitating the tracking of objects within the environment.

    “a finer-grained partitioning of indoor navigable space can significantly contribute in facilitating the tracking of objects within the environment.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationnone given

    The transformation between primal space and dual space in IGNM is bidirectional, so after graph construction nodes are still aware of the topographic space they belong to.

    “Differently from S-MAT and Door-to-Door, the transformation between primal space and dual space is bidirectional: hence, after graph construction, nodes are still aware of the topographic space they belong to.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Premisenone given

    Visibility graph nodes represent the convex corners of a region, including the corners of all obstacles and all portals/entries to the region.

    “Nodes represent the convex corners of a region, including the corners of all obstacles, and additionally all portals/entries to the region.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    Inward offsetting helps in detecting large, open spaces that represent the weak performance of medial axis transform.

    “Inward offsetting helps in detecting large, open spaces that represent the weak performance of medial axis transform.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim.

    Interpretationreasoning

    Creating a visibility-based network as needed using a lightweight mobile device can prove to be a significant challenge.

    “Creating this network as needed, using a lightweight mobile device, can prove to be a significant challenge.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Premisecitation

    The door-to-door approach is organized as a two-level routing strategy: coarse room-to-room connectivity on the floor level to determine direction of movement, and refined use in a single closed space to avoid obstructions and make people transfer in a door-to-door way.

    “The routing strategy is organized as a two-level approach: (1) coarse: it is based on room-to-room connectivity on the floor level to determine the direction of movement and (2) refined: it is used in a single closed space to avoid all kind of obstructions and make people transfer in a door-to-door way.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Liu and Zlatanova 2011

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • It may state two separate claims as one.

    Interpretationnone given

    IGNM preserves the duality relation between primal space (topographic space) and dual space (topology network), while S-MAT and Door-to-Door did not maintain this property.

    “One characteristic of IGNM is that it preserves the duality relation between primal space (i.e., topographic space) and dual space (topology network), while such a property was not maintained either in S-MAT or in Door-to-Door.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Interpretationreasoning

    IGNM improves the quality of localization compared to other network graphs.

    “therefore improving the quality of localization”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Localization is mostly faulty in network graphs generated by both S-MAT and Door-to-Door.

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    The introduction of inward offset polygons can overcome the problems that S-MAT generates in large open spaces.

    “the introduction of the inward offset polygons can overcome the problems that S-MAT generates in large open spaces.”
    Quote in the paperNot a substantial claim
28 of 28

  • The paper relies on this without support.

    The automatic derivation of networks from complex 3D building models is presently one of the most critical obstacles.

    “The automatic derivation of networks from complex 3D building models is presently one of the most critical obstacles.”

  • The paper relies on this without support.

    Door-to-Door path finding is optimal.

    “which is known to be optimal”

  • The paper relies on this without support.

    Localization is mostly faulty in network graphs generated by both S-MAT and Door-to-Door.

    “which is mostly faulty in network graphs generated by both S-MAT and Door-to-Door.”

  • The paper relies on this without support.

    The high number of edges was a disadvantage of Door-to-Door.

    “the high number of edges was a disadvantage of Door-to-Door.”

  • The paper relies on this without support.

    Large open spaces are the main disadvantage of S-MAT.

    “large open spaces that are the main disadvantage of S-MAT”

  • The paper relies on this without support.

    Centerline-based paths are a good representation of natural human behavior within indoor environments.

    “which results to be a good representation of natural human behavior within indoor environments.”
6 of 6

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 3

  1. Kallmann M (2005) Path planning in triangulations. In: Proceedings of the Workshop on Reasoning, Representation, and Learning in Computer Games, International Joint Conference on Artificial Intelligence (IJCAI). Edinburgh, Scotland, pp 49-54
  2. Lee J (2001) 3D data model for representing topological relations of urban features. Paper presented at the 21st Annual ESRI International User Conference, San Diego, CA, 9-13 July 2001
  3. Liu L, Zlatanova S (2011) A “door-to-door” path-finding approach for indoor navigation. In: Proceedings Gi4DM 2011: GeoInformation for Disaster Management. ISPRS, Antalya, Turkey, pp 3–8

Cited by

Sorted by citations

Related

Metadata

Authors
Mortari, F., Clementini, E., Zlatanova, S., & Liu, L.
Affiliations
  • University of L'Aquila (IT)
  • Tongji University (CN)
Corresponding author
Clementini, Eliseo
ORCID
Venue
Applied Geomatics
Publisher
Springer Science and Business Media LLC
Published
2019-06-17
Volume
11
Issue
4
Pages
413-427
ISSN
  • 1866-928X (ISSN-L)
  • 1866-9298
DOI
10.1007/s12518-019-00273-8
OpenAlex
W2950227415
Semantic Scholar
a2d7278c38a6707f99e5aa81b0fe47bea8fc04fb
Cited by
30
References
35
Citations by year
2026: 1 · 2024: 8 · 2023: 4 · 2022: 7 · 2021: 2 · 2020: 8
FWCI
2.80
Type
journal article
Language
en
Open access
Licence
Locations
  • Applied Geomatics (publishedVersion)
  • UNSWorks (University of New South Wales, Sydney, Australia) (submittedVersion, cc-by-nc-nd)
  • Archivio Istituzionale della Ricerca - Università degli Studi dell Aquila (submittedVersion, cc-by-nc-nd)
Funders
University of Cambridge
Topics
  • Physical Sciences › Engineering › Building and Construction › 3D Modeling in Geospatial Applications
  • Social Sciences › Urban Studies › Urban Design and Spatial Analysis
  • Social Sciences › Geography, Planning and Development › Geographic Information Systems Studies
Keywords
indoor navigation, navigation model, IndoorGML, CityGML, network extraction, medial axis, visibility graph, 3D building modeling, human navigation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-21, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-21 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review