Fengyan Lun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Fengyan Lun has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
Chuang Wang
1 shared publication- Lidong Gu
School of Mechanical and Electrical Engineering, Changchun University of Science and Technology, Changchun 130022, China
1 shared publication - Meichao Xu
School of Mechanical and Electrical Engineering, ChangChun University of Science and Technology, China
1 shared publication - Renjun Li
Changchun University of Science and Technology
1 shared publication Xiaolong Zhou
1 shared publication- Yiwen Zhang
School of Mechanical and Electrical Engineering, Changchun University of Science and Technology, Changchun 130022, China
1 shared publication
Latest Publications
- Open Access
- 28 February 2026
- Published by Elsevier in SSRN Electronic Journal
Research on a multi-scale tomato ripeness detection method based on an enhanced YOLOv13s
This study aims to enhance tomato ripeness detection for harvesting robots by proposing an improved YOLOv13 algorithm named YOLOv13-SSW. To increase accuracy and efficiency, three key enhancements were made: The SimAM attention mechanism was integrated into the backbone to adaptively highlight salient features. The original DS-C3k2 convolution modules were replaced with DS-C3k2_SAConv modules, utilizing SAConv to reduce spatial and channel redundancy for faster detection and fewer parameters The CIoU loss function was replaced with WIoUv3 to improve convergence. Evaluated on 5,000 images over 300 epochs, the model achieved 91.612% precision, 89.678% recall, and 92.319% mAP. Compared to the baseline YOLOv13s, this represents significant improvements of 5.553%, 4.62%, and 8.033%, respectively. Comparative and ablation experiments confirmed the model's strong generalization, robustness, and the necessity of each improvement, providing a valuable reference for similar agricultural vision tasks.
- 28 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessResearch on a multi-scale tomato ripeness detection method based on an enhanced YOLOv13s
This study aims to enhance tomato ripeness detection for harvesting robots by proposing an improved YOLOv13 algorithm named YOLOv13-SSW. To increase accuracy and efficiency, three key enhancements were made: The SimAM attention mechanism was integrated into the backbone to adaptively highlight salient features. The original DS-C3k2 convolution modules were replaced with DS-C3k2_SAConv modules, utilizing SAConv to reduce spatial and channel redundancy for faster detection and fewer parameters The CIoU loss function was replaced with WIoUv3 to improve convergence. Evaluated on 5,000 images over 300 epochs, the model achieved 91.612% precision, 89.678% recall, and 92.319% mAP. Compared to the baseline YOLOv13s, this represents significant improvements of 5.553%, 4.62%, and 8.033%, respectively. Comparative and ablation experiments confirmed the model's strong generalization, robustness, and the necessity of each improvement, providing a valuable reference for similar agricultural vision tasks.
- Open Access
- 26 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 432-436
Knowledge Graph and Mind Mappings of Mechanisms and Machine Theory and Their Applications in Digital Education
This study addressed the issues in the course of Mechanisms and Machine Theory, such as scattered knowledge points and lack of systematic organization, in engineering universities. We applied knowledge graph and Mind mapping techniques to organize and reconstruct the knowledge points, combining with the Rain Classroom application, to explore the construction of digital education ecology. Through Pre-class, In-class and Post-class teaching practices, the results show that the combination of knowledge graph, Mind mappings and Rain Classroom not only solves the problem of students' inability to understand the knowledge system of the course fully and effectively but also promotes lecturers' understanding and control of the overall curriculum, which can provide new ideas for the construction of digital education ecology in engineering universities.
- 26 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 432-436
Open AccessKnowledge Graph and Mind Mappings of Mechanisms and Machine Theory and Their Applications in Digital Education
This study addressed the issues in the course of Mechanisms and Machine Theory, such as scattered knowledge points and lack of systematic organization, in engineering universities. We applied knowledge graph and Mind mapping techniques to organize and reconstruct the knowledge points, combining with the Rain Classroom application, to explore the construction of digital education ecology. Through Pre-class, In-class and Post-class teaching practices, the results show that the combination of knowledge graph, Mind mappings and Rain Classroom not only solves the problem of students' inability to understand the knowledge system of the course fully and effectively but also promotes lecturers' understanding and control of the overall curriculum, which can provide new ideas for the construction of digital education ecology in engineering universities.