Stuardo Lucho
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Stuardo Lucho has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Sylvie Treuillet
PRISME, Université d'Orléans, Orleans, France
3 shared publications - Abraham Dávila
Pontifical Catholic University of Peru
2 shared publications - Karin Melendez
Pontificia Universidad Católica del Perú
2 shared publications - Xavier Brunetaud
University of Orléans
2 shared publications - Xavier Desquesnes
PRISME, Université d'Orléans, Orleans, France
2 shared publications Angelo Velarde
1 shared publication
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 13-21
Spalling segmentation on tuffeau limestone heritage facades using RGB–thermal fusion
Spalling is a common deterioration process in the porous tuffeau limestone of Loire Valley heritage monuments. It involves the local detachment of surface layers and, when it progresses, leaves the underlying stone exposed to further weathering. At present, damaged areas are generally mapped through visual inspection by conservation specialists. Although this procedure provides valuable expert assessment, applying it systematically to large façades is time-consuming, and the resulting maps may be difficult to reproduce across inspection campaigns. This paper investigates the automatic segmentation of spalling by combining RGB images and passive infrared thermography (IRT), using data acquired on the south-east façade of the Château de Chambord. We introduce an RGB–thermal dataset covering ten spalling defects, with 73 registered image pairs and geometrically consistent masks generated through 3D ortho-reprojection across viewpoints and sensors. We benchmark single-modality baselines and early, intermediate, and late RGB–thermal fusion strategies across several backbones. We further propose TTV-Net, a thermo-visual segmentation architecture that injects thermal evidence into RGB features through context-gated thermal injection based on normalized temperature, local thermal residuals, and thermal-gradient cues. Under grouped three-fold cross-validation by defect, TTV-Net obtains the highest mean performance, reaching an IoU of 0.818 and an F1-score of 0.899. It improves slightly over the strongest standard fusion baseline, which reaches 0.795 IoU, and more clearly over the best RGB-only and thermal-only models, which reach 0.705 and 0.482 IoU, respectively.
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 13-21
Open AccessSpalling segmentation on tuffeau limestone heritage facades using RGB–thermal fusion
Spalling is a common deterioration process in the porous tuffeau limestone of Loire Valley heritage monuments. It involves the local detachment of surface layers and, when it progresses, leaves the underlying stone exposed to further weathering. At present, damaged areas are generally mapped through visual inspection by conservation specialists. Although this procedure provides valuable expert assessment, applying it systematically to large façades is time-consuming, and the resulting maps may be difficult to reproduce across inspection campaigns. This paper investigates the automatic segmentation of spalling by combining RGB images and passive infrared thermography (IRT), using data acquired on the south-east façade of the Château de Chambord. We introduce an RGB–thermal dataset covering ten spalling defects, with 73 registered image pairs and geometrically consistent masks generated through 3D ortho-reprojection across viewpoints and sensors. We benchmark single-modality baselines and early, intermediate, and late RGB–thermal fusion strategies across several backbones. We further propose TTV-Net, a thermo-visual segmentation architecture that injects thermal evidence into RGB features through context-gated thermal injection based on normalized temperature, local thermal residuals, and thermal-gradient cues. Under grouped three-fold cross-validation by defect, TTV-Net obtains the highest mean performance, reaching an IoU of 0.818 and an F1-score of 0.899. It improves slightly over the strongest standard fusion baseline, which reaches 0.795 IoU, and more clearly over the best RGB-only and thermal-only models, which reach 0.705 and 0.482 IoU, respectively.
- 3 August 2026
- Published by Springer Nature
- p. 576-590
Gradient Consistency Focal Dice Loss and Multi-scale Attention for Accurate Segmentation of Mortar Joints in Stone Masonry
- 3 August 2026
- Published by Springer Nature
- p. 576-590
Gradient Consistency Focal Dice Loss and Multi-scale Attention for Accurate Segmentation of Mortar Joints in Stone Masonry
- Open Access
- 19 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 173824-173835
Can Deep Learning Wound Segmentation Algorithms Developed for a Dataset Be Effective for Another Dataset? A Specific Focus on Diabetic Foot Ulcers
Diabetic foot ulcers (DFU) represent a severe complication, often resulting from poor glycemic control, neuropathy, peripheral vascular disease, or inadequate foot care. DFUs can lead to significant morbidity, including amputation and, in severe cases, can be fatal. Recently, advancements in computer vision technologies based on artificial intelligence (AI) have shown promise in DFU management. Particularly deep learning (DL) models such as U-Net and other models and techniques, were utilized to enhance wound segmentation accuracy. This research focuses on evaluating the generalization capabilities of DL models across different DFU datasets. Specifically, we investigated whether models trained on one dataset can be effective when utilized on another dataset, addressing the challenge of cross-dataset generalization. We employed 7 popular DL models, U-Net-VGG16, U-Net-EfficientNetV2S, ABANet, Ma-Net, LinkNet, DeepLabV3+, and Segment Anything Model (SAM), with 2 DFU datasets: FUSeg challenge and DFUC challenge. A total of 54 experiments were conducted plus 27 for SAM, involving training on one dataset, and testing on another, as well as training and testing on combined datasets. The results indicate substantial variability in segmentation performance when models trained on one dataset are tested on another, highlighting the influence of dataset characteristics on model generalization. The study underscores the importance of using diverse and comprehensive datasets to develop robust DL models for DFU segmentation and its generalization. This research contributes to the understanding of DL model performance in medical image segmentation and emphasizes the need for standardized datasets in improving DFU management through computer vision.
- 19 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 173824-173835
Open AccessCan Deep Learning Wound Segmentation Algorithms Developed for a Dataset Be Effective for Another Dataset? A Specific Focus on Diabetic Foot Ulcers
Diabetic foot ulcers (DFU) represent a severe complication, often resulting from poor glycemic control, neuropathy, peripheral vascular disease, or inadequate foot care. DFUs can lead to significant morbidity, including amputation and, in severe cases, can be fatal. Recently, advancements in computer vision technologies based on artificial intelligence (AI) have shown promise in DFU management. Particularly deep learning (DL) models such as U-Net and other models and techniques, were utilized to enhance wound segmentation accuracy. This research focuses on evaluating the generalization capabilities of DL models across different DFU datasets. Specifically, we investigated whether models trained on one dataset can be effective when utilized on another dataset, addressing the challenge of cross-dataset generalization. We employed 7 popular DL models, U-Net-VGG16, U-Net-EfficientNetV2S, ABANet, Ma-Net, LinkNet, DeepLabV3+, and Segment Anything Model (SAM), with 2 DFU datasets: FUSeg challenge and DFUC challenge. A total of 54 experiments were conducted plus 27 for SAM, involving training on one dataset, and testing on another, as well as training and testing on combined datasets. The results indicate substantial variability in segmentation performance when models trained on one dataset are tested on another, highlighting the influence of dataset characteristics on model generalization. The study underscores the importance of using diverse and comprehensive datasets to develop robust DL models for DFU segmentation and its generalization. This research contributes to the understanding of DL model performance in medical image segmentation and emphasizes the need for standardized datasets in improving DFU management through computer vision.
- Open Access
- 19 June 2024
- Published by MDPI AG in Journal of Imaging
- Vol. 10 (6) , 148
Weakly Supervised SVM-Enhanced SAM Pipeline for Stone-by-Stone Segmentation of the Masonry of the Loire Valley Castles
The preservation of historical monuments presents a formidable challenge, particularly in monitoring the deterioration of building materials over time. Chateau de Chambord’s facade suffers from common issues such as flaking and spalling, which require meticulous stone and joint mapping from experts manually for restoration efforts. Advancements in computer vision have allowed machine-learning models to help in the automatic segmentation process. In this research, a custom architecture defined as SAM-SVM is proposed, to perform stone segmentation, based on the Segment Anything Model (SAM) and Support Vector Machines (SVM). By exploiting the zero-shot learning capabilities of SAM and its customizable input parameters, we obtain segmentation mask for stones and joints, which are then classified using SVM. Two more SAMs (three in total) are used, depending on how many stones are left to segment. Through extensive experimentation and evaluation, supported by computer vision methods, the proposed architecture achieves a Dice coefficient of 85%. Our results highlight the potential of SAM in cultural heritage conservation, providing a scalable and efficient solution for stone segmentation in historic monuments. This research contributes valuable insights and methodologies to the ongoing conservation efforts of Château de Chambord and could be extrapolated to other monuments.
- 19 June 2024
- Published by MDPI AG in Journal of Imaging
- Vol. 10 (6) , 148
Open AccessWeakly Supervised SVM-Enhanced SAM Pipeline for Stone-by-Stone Segmentation of the Masonry of the Loire Valley Castles
The preservation of historical monuments presents a formidable challenge, particularly in monitoring the deterioration of building materials over time. Chateau de Chambord’s facade suffers from common issues such as flaking and spalling, which require meticulous stone and joint mapping from experts manually for restoration efforts. Advancements in computer vision have allowed machine-learning models to help in the automatic segmentation process. In this research, a custom architecture defined as SAM-SVM is proposed, to perform stone segmentation, based on the Segment Anything Model (SAM) and Support Vector Machines (SVM). By exploiting the zero-shot learning capabilities of SAM and its customizable input parameters, we obtain segmentation mask for stones and joints, which are then classified using SVM. Two more SAMs (three in total) are used, depending on how many stones are left to segment. Through extensive experimentation and evaluation, supported by computer vision methods, the proposed architecture achieves a Dice coefficient of 85%. Our results highlight the potential of SAM in cultural heritage conservation, providing a scalable and efficient solution for stone segmentation in historic monuments. This research contributes valuable insights and methodologies to the ongoing conservation efforts of Château de Chambord and could be extrapolated to other monuments.
- Open Access
- 18 March 2018
- Published by AISTI in RISTI - Revista Ibérica de Sistemas e Tecnologias de Informação
- Vol. 26 (26) , 1-16
Influencia de Factores de Entorno en la Implementación del estándar ISO/IEC 29110
El estándar ISO/IEC 29110 es usado por algunas empresas peruanas con el fin de mejorar la calidad del software que elaboran y ser más productivos. Sin embargo, existen factores que influyen en el equipo durante la mejora de procesos de software, por lo cual, identificar adecuadamente estos factores, permitirá mejorar el nivel de adopción en futuras implementaciones. El objetivo de este estudio es analizar los factores del entorno que influyen en la adopción del estándar ISO/IEC 29110, en su perfil básico, en 4 pequeñas empresas desarrolladoras de software, utilizando la metodología de estudio de caso múltiple. Cada empresa trabajó bajo su propio contexto y definió sus propios objetivos de mejora. Del conjunto de factores del entorno analizados, se encontró como resultado que 5 de ellos influenciaron positivamente en todas las empresas sin importar el contexto, 2 ellos negativamente y 2 de ellos tuvieron relación entre factores debido al contexto
- 18 March 2018
- Published by AISTI in RISTI - Revista Ibérica de Sistemas e Tecnologias de Informação
- Vol. 26 (26) , 1-16
Open AccessInfluencia de Factores de Entorno en la Implementación del estándar ISO/IEC 29110
El estándar ISO/IEC 29110 es usado por algunas empresas peruanas con el fin de mejorar la calidad del software que elaboran y ser más productivos. Sin embargo, existen factores que influyen en el equipo durante la mejora de procesos de software, por lo cual, identificar adecuadamente estos factores, permitirá mejorar el nivel de adopción en futuras implementaciones. El objetivo de este estudio es analizar los factores del entorno que influyen en la adopción del estándar ISO/IEC 29110, en su perfil básico, en 4 pequeñas empresas desarrolladoras de software, utilizando la metodología de estudio de caso múltiple. Cada empresa trabajó bajo su propio contexto y definió sus propios objetivos de mejora. Del conjunto de factores del entorno analizados, se encontró como resultado que 5 de ellos influenciaron positivamente en todas las empresas sin importar el contexto, 2 ellos negativamente y 2 de ellos tuvieron relación entre factores debido al contexto
- 12 October 2017
- Published by Springer Nature
- p. 82-93
Analysis of environmental factors in the adoption of ISO/IEC 29110. Multiple case study
- 12 October 2017
- Published by Springer Nature
- p. 82-93
Analysis of environmental factors in the adoption of ISO/IEC 29110. Multiple case study