Skip to content
DraconiticResearch
Search Sources API
190,851 studies31 sourcesUpdated 53 min ago
Peer-reviewedCC BYScientific Reports

Edge context fusion transformer for accurate and efficient photovoltaic electroluminescence defect detection

  • Xing Liu1
  • Rong Zhu2
  • Hao Yan3
  1. 1School of Computer Science, Shandong Xiehe University
  2. 2School of Modern Information Industry, Guangzhou College of Commerce
  3. 3Faculty of Data Science, City University of Macau
Read full textPDFDOICite

Abstract

Photovoltaic electroluminescence (EL) imaging enables nondestructive identification of latent module defects, but automatic detection remains challenging because many defects are small, low-contrast, boundary-ambiguous, and distributed across multiple scales. This study proposes ECF-DETR, an Edge Context Fusion Detection Transformer built on RT-DETR-R18. Boundary-Aware Multi-Scale Edge Enhancement is designed to retain weak boundary cues, Bidirectional Semantic-Spatial Cross-Scale Fusion facilitates information exchange between adjacent feature levels, and Gated Multi-Branch Context Aggregation refines nonlinear contextual representation after cross-scale fusion. On PVEL-AD, ECF-DETR achieves 94.5% [email protected] and 66.0% [email protected]:0.95, improving RT-DETR-R18 by 3.1 and 2.2 percentage points while reducing the parameter count from 19.88 to 16.33 M. It requires 57.2 GFLOPs and reaches 92 FPS on an NVIDIA GeForce RTX 4090D at an input resolution of 640 × 640 pixels. On PV-Multi-Defect, it obtains 87.6% [email protected] and 58.5% [email protected]:0.95. TensorRT FP16 deployment on an NVIDIA Jetson Orin Nano with 8 GB of shared memory achieves 42 FPS, compared with 40 FPS for RT-DETR-R18 and 18 FPS for RT-DETR-R50. Under the evaluated datasets and hardware settings, these results indicate a favorable accuracy–efficiency balance for photovoltaic EL defect detection.

Related

  1. Cracks-YOLO: an improved YOLOv5 tongue crack detection method based on switchable atrous convolution and transformer
    Miaomiao Ding, Meiyi Wu, Bochen Shen +2
    Frontiers in medicinePeer-reviewedCC BYFull textGeneral medicineEurope PMC
  2. HayabusaNet: Hybrid attention-based multiscale fusion CNN for accurate and efficient brain tumor classification in MRI scans
    Rizal Dwi Prayogo, Siti Amatullah Karimah, Hidetaka Nambo
    Expert Systems with ApplicationsPeer-reviewedCC BY5 citedNeurologyOpenAlex
  3. An efficient brain tumor classification system with accurate recognition and complementary downsampling blocks
    Jincan Zhang, Chuanqi Cai, XingHua Tan +4
    PLOS OnePeer-reviewedCC BY 4.0Full textMedical imaging & radiologyOncologyPLOS
  4. EPRA U-Net: an efficient pyramid residual attention framework for accurate infarct segmentation in diffusion-weighted MRI.
    Hasan Ulutas, Muhammet Emin Sahin, Mustafa Fatih Erkoc +4
    BMC medical imagingPeer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMCDOAJ
  5. Edge efficient human activity recognition using a quantized patch based transformer
    Aasif Rashid Khanday, Rajendra Kumar, Yonis Gulzar +1
    Scientific ReportsPeer-reviewedCC BYFull textGeneral medicineDOAJ

Cite

BibTeXRIS
BibTeX
@article{liu2026edge,
  title = {{Edge context fusion transformer for accurate and efficient photovoltaic electroluminescence defect detection}},
  author = {Xing Liu and Rong Zhu and Hao Yan},
  journal = {Scientific Reports},
  year = {2026},
  month = sep,
  doi = {10.1038/s41598-026-70180-7},
  url = {https://doi.org/10.1038/s41598-026-70180-7},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Edge context fusion transformer for accurate and efficient photovoltaic electroluminescence defect detection
AU  - Xing Liu
AU  - Rong Zhu
AU  - Hao Yan
PY  - 2026
DA  - 2026/09/01
JO  - Scientific Reports
DO  - 10.1038/s41598-026-70180-7
UR  - https://doi.org/10.1038/s41598-026-70180-7
AB  - Photovoltaic electroluminescence (EL) imaging enables nondestructive identification of latent module defects, but automatic detection remains challenging because many defects are small, low-contrast, boundary-ambiguous, and distributed across multiple scales. This study proposes ECF-DETR, an Edge Context Fusion Detection Transformer built on RT-DETR-R18. Boundary-Aware Multi-Scale Edge Enhancement is designed to retain weak boundary cues, Bidirectional Semantic-Spatial Cross-Scale Fusion facilitates information exchange between adjacent feature levels, and Gated Multi-Branch Context Aggregation refines nonlinear contextual representation after cross-scale fusion. On PVEL-AD, ECF-DETR achieves 94.5% [email protected] and 66.0% [email protected]:0.95, improving RT-DETR-R18 by 3.1 and 2.2 percentage points while reducing the parameter count from 19.88 to 16.33 M. It requires 57.2 GFLOPs and reaches 92 FPS on an NVIDIA GeForce RTX 4090D at an input resolution of 640 × 640 pixels. On PV-Multi-Defect, it obtains 87.6% [email protected] and 58.5% [email protected]:0.95. TensorRT FP16 deployment on an NVIDIA Jetson Orin Nano with 8 GB of shared memory achieves 42 FPS, compared with 40 FPS for RT-DETR-R18 and 18 FPS for RT-DETR-R50. Under the evaluated datasets and hardware settings, these results indicate a favorable accuracy–efficiency balance for photovoltaic EL defect detection.
SN  - 2045-2322
ER  -