Ramirez Agudelo, Oscar Hernan und Shewatkar, Akshay Narendra und Milana, Edoardo und Aydin, Roland und Franke, Kai (2023) Enhancing the quality of gauge images captured in haze and smoke scenes through deep learning. In: Proceedings of SPIE - The International Society for Optical Engineering, 12675, Seiten 68-84. SPIE Digital Library. SPIE Optical Engineering + Applications 2023, 2023-08-20 - 2023-08-24, San Diego, California, USA. doi: 10.1117/12.2679809. ISSN 0277-786X.
![]() |
PDF
46MB |
Offizielle URL: https://doi.org/10.1117/12.2679809
Kurzfassung
Images captured in hazy and smoky environments suffer from reduced visibility, posing a challenge when monitoring infrastructures and hindering emergency services during critical situations. The proposed work investigates the use of deep learning models to enhance the automatic, machine-based readability of gauge in smoky environments, with accurate gauge data interpretation serving as a valuable tool for first responders. The study utilizes two deep learning architectures, FFA-Net and AECR-Net, to improve the visibility of gauge images, corrupted with light up to dense haze and smoke. Since benchmark datasets of analog gauge images were unavailable, two synthetic datasets are generated using the Unreal Engine: a synthetic haze (approx. 4800 images) and a synthetic smoke (approx. 9600 images). Two datasets and two deep-learning frameworks allow the investigation of four different models. The models are trained with an 80% train, 10% validation, and 10% test split for the haze and smoke dataset, respectively. As a result, more robust results are retrieved from the AECR-Net, when compared to FFA-Net and to a prior-based model. For instance, for the synthetic haze dataset, the SSIM and PSNR metrics are about 0.98 and 43 dB obtained with AECR-Net, respectively, comparing well to state-of-the art results. The results from the synthetic smoke dataset are poorer, however the trained models still achieve interesting results. In general, imaging in the presence of smoke are more difficult to enhance given the inhomogeneity and high density. Secondly, FFA-Net and AECR-Net are implemented to dehaze and not to desmoke images. This work shows that use of deep learning architectures can improve the quality of analog gauge images captured in smoke and haze scenes immensely. Finally, the enhanced output images can be successfully post-processed for automatic autonomous reading of gauges.
elib-URL des Eintrags: | https://elib.dlr.de/198015/ | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Dokumentart: | Konferenzbeitrag (Vortrag) | ||||||||||||||||||||||||
Titel: | Enhancing the quality of gauge images captured in haze and smoke scenes through deep learning | ||||||||||||||||||||||||
Autoren: |
| ||||||||||||||||||||||||
Datum: | 4 Oktober 2023 | ||||||||||||||||||||||||
Erschienen in: | Proceedings of SPIE - The International Society for Optical Engineering | ||||||||||||||||||||||||
Referierte Publikation: | Ja | ||||||||||||||||||||||||
Open Access: | Ja | ||||||||||||||||||||||||
Gold Open Access: | Nein | ||||||||||||||||||||||||
In SCOPUS: | Ja | ||||||||||||||||||||||||
In ISI Web of Science: | Ja | ||||||||||||||||||||||||
Band: | 12675 | ||||||||||||||||||||||||
DOI: | 10.1117/12.2679809 | ||||||||||||||||||||||||
Seitenbereich: | Seiten 68-84 | ||||||||||||||||||||||||
Herausgeber: |
| ||||||||||||||||||||||||
Verlag: | SPIE Digital Library | ||||||||||||||||||||||||
Name der Reihe: | Applications of Machine Learning 2023 | ||||||||||||||||||||||||
ISSN: | 0277-786X | ||||||||||||||||||||||||
Status: | veröffentlicht | ||||||||||||||||||||||||
Stichwörter: | gauge reader, synthetic data, image enhancement, unreal engine, safety and security, dehaze and desmoke, deep learning | ||||||||||||||||||||||||
Veranstaltungstitel: | SPIE Optical Engineering + Applications 2023 | ||||||||||||||||||||||||
Veranstaltungsort: | San Diego, California, USA | ||||||||||||||||||||||||
Veranstaltungsart: | internationale Konferenz | ||||||||||||||||||||||||
Veranstaltungsbeginn: | 20 August 2023 | ||||||||||||||||||||||||
Veranstaltungsende: | 24 August 2023 | ||||||||||||||||||||||||
Veranstalter : | SPIE Optics + Photonics | ||||||||||||||||||||||||
HGF - Forschungsbereich: | keine Zuordnung | ||||||||||||||||||||||||
HGF - Programm: | keine Zuordnung | ||||||||||||||||||||||||
HGF - Programmthema: | keine Zuordnung | ||||||||||||||||||||||||
DLR - Schwerpunkt: | keine Zuordnung | ||||||||||||||||||||||||
DLR - Forschungsgebiet: | keine Zuordnung | ||||||||||||||||||||||||
DLR - Teilgebiet (Projekt, Vorhaben): | keine Zuordnung | ||||||||||||||||||||||||
Standort: | Rhein-Sieg-Kreis | ||||||||||||||||||||||||
Institute & Einrichtungen: | Institut für den Schutz terrestrischer Infrastrukturen Institut für den Schutz terrestrischer Infrastrukturen > Detektionssysteme | ||||||||||||||||||||||||
Hinterlegt von: | Ramirez Agudelo, Oscar Hernan | ||||||||||||||||||||||||
Hinterlegt am: | 12 Okt 2023 16:38 | ||||||||||||||||||||||||
Letzte Änderung: | 24 Apr 2024 20:58 |
Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags