EUVIMEDEUVIMED entdecken
ÜbersichtsarbeitOpen Access · CC BY 4.0

3D Anomaly Detection for Sensor-Based Industrial Inspection: A Critical Survey of Methods, Benchmarks, and Deployment Constraints

Lars Fichtel iD, Radu Timofte iD, Dominik Seuß iD

Sensors · 23.10.2026 · 26 (21) · 6717

DOI: 10.3390/s26216717 ↗

Kurz gefasst

Übersichtsarbeit in der Zeitschrift Sensors, online erschienen am 23.10.2026. Verfasst von Lars Fichtel und 2 weiteren Autor:innen. Beteiligte Einrichtung: Center for Artificial Intelligence (CAIRO), Technical University of Applied Sciences Würzburg-Schweinfurt, Münzstraße 12, 97070 Würzburg, Germany. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 186 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Alexander von Humboldt Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Some industrial defects are geometric and invisible to a color camera, and anomaly detection on point clouds, depth maps and RGB-D images grew up around them. This survey covers 2020 to 21 August 2026, 94 detectors and 17 benchmark datasets, with an executable search protocol tested against the corpus and every collated number attributed to its source. Methods sit on two axes: six mutually exclusive scoring mechanisms (feature and memory matching, reconstruction, teacher–student discrepancy, density estimation, registration, and foundation-model transfer) and eight separable design dimensions, so fusion and zero-shot supervision are properties, not families. Benchmark results are collated with protocol, metric variant and peer-review status rather than ranked: on MVTec 3D-AD, the highest image-level area under the receiver operating characteristic curve (I-AUROC) is 98.4% peer-reviewed and 97.2% in a preprint, under differing protocols; the five highest span 1.1 points. Color-and-geometry fusion leads where both are captured; density estimation and reconstruction lead on point-cloud-only, few-shot Real3D-AD. Few of the 108 works examined report enough of the fields a runtime comparison needs. Sensing robustness is mostly tested along conditions an ablation already varies, rarely along those an instrument would introduce. A closing guide states what the evidence lets a practitioner select.

Schlagwörter

Stichwörter der Autor:innen

3D anomaly detection3D sensingRGB-D sensingbenchmark datasetsfew-shot learningindustrial inspectionmultimodal sensor fusionpoint cloudsunsupervised learningzero-shot learning

Autor:innen und Institutionen

  • Center for Artificial Intelligence (CAIRO), Technical University of Applied Sciences Würzburg-Schweinfurt, Münzstraße 12, 97070 Würzburg, Germany.
  • Computer Vision Laboratory, University of Würzburg, 97074 Würzburg, Germany.
  • Fraunhofer Institute for Integrated Circuits IIS, 91058 Erlangen, Germany.

ORCID-Profile: Lars Fichtel, Radu Timofte, Dominik Seuß

Förderung

  • Alexander von Humboldt Foundation

Zitieren

APA

Fichtel, L., Timofte, R., & Seuß, D. (2026). 3D Anomaly Detection for Sensor-Based Industrial Inspection: A Critical Survey of Methods, Benchmarks, and Deployment Constraints. Sensors, 26(21), 6717. https://doi.org/10.3390/s26216717

Vancouver

Fichtel L, Timofte R, Seuß D. 3D Anomaly Detection for Sensor-Based Industrial Inspection: A Critical Survey of Methods, Benchmarks, and Deployment Constraints. Sensors. 2026;26(21):6717. doi:10.3390/s26216717

BibTeX
@article{Fichtel2026,
  author = {Fichtel, Lars and Timofte, Radu and Seuß, Dominik},
  title = {3D Anomaly Detection for Sensor-Based Industrial Inspection: A Critical Survey of Methods, Benchmarks, and Deployment Constraints},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {6717},
  doi = {10.3390/s26216717}
}
Kostenlos starten →