Zoubir BARRAZ

Mr. Zoubir BARRAZ

Photogrammetry-Cartography Department, Hassan II Institute of Agronomy & Veterinary Medicine, Rabat, Rabat-Salé-Kénitra, Morocco
orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 March 2025 in Technologies
Verified
Journal Article
Open access
PDF file
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
ISImane Sebari
Imane Sebari
KAKenza AitElKadi
Kenza AitElKadi
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula
This paper provides an in-depth literature review on image processing techniques, focusing on deep learning approaches for anomaly detection and classification in photovoltaics. It examines key components of UAV-based PV inspection, including data acquisition protocols, panel segmentation and geolocation, anomaly classification, and optimizations for model generalization. Furthermore, challenges related to domain adaptation, dataset limitations, and multimodal fusion of RGB and thermal data are also discussed. Finally, research gaps and opportunities are analyzed to create a holistic, scalable, and real-time inspection workflow for large-scale installation. This review serves as a reference for researchers and industry professionals to advance UAV-based PV inspection.

1 January 2025 in Results in Engineering
Verified
Journal Article
Open access
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
ISImane Sebari
Imane Sebari
NLNassim Lamrini
Nassim Lamrini
KKKenza Ait El Kadi
Kenza Ait El Kadi
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula

9 May 2025 in Energy and AI
Verified
Journal Article
Open access
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
ISImane Sebari
Imane Sebari
HOHicham Oufettoul
Hicham Oufettoul
KKKenza Ait el Kadi
Kenza Ait el Kadi
NLNassim Lamrini
Nassim Lamrini
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula

15 July 2025 in Cleaner Engineering and Technology
Verified
Journal Article
Open access
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
ISImane Sebari
Imane Sebari
NLNassim Lamrini
Nassim Lamrini
KKKenza Ait El Kadi
Kenza Ait El Kadi
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula

8 May 2023
Verified
Conference Paper
Closed access
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula
NLNassim Lamrini
Nassim Lamrini
PSPr. Imane Sebari
Pr. Imane Sebari
Anomaly detection in photovoltaic systems using deep learning models can help further automatize classification tasks and support the operator in decision making. However, the performance of pretrained deep learning models decreases in production owing to the difference between training and production data (domain shift). This performance gap can lead to serious problems when monitoring photovoltaic (PV) farms. To this end, this study proposes an approach to adapt new instances to fit the constraints of the input layer of the pretrained model by image segmentation using Niblack thresholding, padding, and resizing. In addition, the localization of single panels was calculated using metadata. Moreover, a retraining cycle was proposed to include new instances in the training process to fine-tune the models and adapt the weights to the images taken during production. The proposed algorithm extracted 81% of the panels from the images fed into it. However, the pretrained model, which was used to classify the segmented images to test its performance, did not perform well. This indicates the importance of the retraining process for optimizing the model for production.

1 July 2024
Verified
Conference Paper
Closed access
Zoubir BARRAZZoubir BARRAZ
Zoubir BARRAZ
ISImane Sebari
Imane Sebari
KKKenza Ait El Kadi
Kenza Ait El Kadi
NLNassim Lamrini
Nassim Lamrini
IAIbtihal Ait Abdelmoula
Ibtihal Ait Abdelmoula
Using trained deep learning models on an out-of-distribution dataset could affect those models’ performance, which is most likely to decline due to the domain shift between the training domain and the target domain. To this end, Domain Generalization (DG) is one of the tools used to improve the generalization rate of the model and increase its adaptability to unseen data. Using a data-based domain generalization, this study intends to create a solar module segmentation model using thermal UAV images, by including various type of technologies and representations of solar modules in training data. Developing the generalization of the solar panel segmentation could enhance the performance in the workflow of automatic photovoltaic plants monitoring. The thermal images dataset was collected from six photovoltaic sites in Benguerir - Morocco in various shooting conditions. The global capacity of these systems is 1MW and they are equipped with different technologies. The training of this model achieved 99% mAP@0.5 for testing and could perform well on unseen data, that is different from the training data, ensuring the effectiveness of the approach to generalize the trained Deep learning models which can be used in other domains.