Deep learning‐based automatic detection of multitype defects in photovoltaic modules and application in real production line
Bibliographic record
Abstract
Abstract Automatic defect detection in electroluminescence (EL) images of photovoltaic (PV) modules in production line remains as a challenge to replace time‐consuming and expensive human inspection and improve capacity. This paper presents a deep learning‐based automatic detection of multitype defects to fulfill inspection requirements of production line. At first, a database composed of 5983 labeled EL images of defective PV modules is built, and 19 types of identified defects are introduced. Next, a convolutional neural network is trained on top‐14 defects, and the best model is selected and tested, achieving 70.2% mAP50 (mean average precision with at least 50% localization accuracy). Then, through analyzing an object detection‐based confusion matrix, recognition bias and detection compensation in missed defects that restrain the best model's mAP50 are discovered to be harmless to normal/defective module classification in real production line. Finally, after setting specific screen criteria for different types of defects, normal/defective module classification is conducted on additionally collected 4791 EL images of PV modules on 3 days, and the best model achieves balanced scores of 95.1%, 96.0%, and 97.3%, respectively. As a result, this method surely has a highly promising potential to be adopted in real production line.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".