Research on Surface Defect Detection Algorithm of Lithium Battery Based on Multi-Modal Deep Learning
Journal of Engineering System · 2026-03 · 4 (1) · 29-39
DOI: 10.62517/jes.202602105 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Engineering System, 2026-03. Verfasst von Yuze Yang. Die Arbeit nennt 53 Literaturangaben.
Abstract (Originalsprache)
With the rapid advancement of new energy technologies, lithium-ion batteries as core energy storage components have been widely adopted in electric vehicles, portable devices, and energy storage systems, becoming a vital component of modern energy solutions. However, surface defects in lithium-ion batteries significantly compromise their performance, lifespan, and safety-particularly under high-energy-density conditions and frequent usage scenarios where these defects may trigger short circuits, overheating, or even explosions. Current traditional detection methods for lithium-ion battery surface defects primarily rely on manual inspection or conventional computer vision technologies. These approaches demonstrate notable limitations in precision and efficiency, especially when handling various defect types and performing complex defect detection in intricate scenarios. To address the challenges in surface defect detection of lithium batteries, this study proposes an intelligent detection algorithm based on multimodal deep learning. We have designed and implemented an innovative model framework that combines image processing with text analysis to comprehensively improve the accuracy and robustness of defect detection. This method utilizes deep learning models to extract visual features of surface defects and incorporates defect-related textual descriptions as auxiliary information, significantly enhancing the precision in identifying and localizing various types of defects. Experimental results show that the proposed detection algorithm achieves an accuracy of 95.6%, representing a significant performance improvement over existing methods, especially demonstrating strong adaptability and generalization capability in complex scenarios. This research not only provides an efficient and precise technical solution for defect detection in lithium battery production but also offers important theoretical and methodological support for other industrial defect detection tasks.
Zitieren
APA
Yang, Y. (2026). Research on Surface Defect Detection Algorithm of Lithium Battery Based on Multi-Modal Deep Learning. Journal of Engineering System, 4(1), 29-39. https://doi.org/10.62517/jes.202602105
Vancouver
Yang Y. Research on Surface Defect Detection Algorithm of Lithium Battery Based on Multi-Modal Deep Learning. Journal of Engineering System. 2026;4(1):29-39. doi:10.62517/jes.202602105
BibTeX
@article{Yang2026,
author = {Yang, Yuze},
title = {Research on Surface Defect Detection Algorithm of Lithium Battery Based on Multi-Modal Deep Learning},
journal = {Journal of Engineering System},
year = {2026},
volume = {4},
number = {1},
pages = {29--39},
doi = {10.62517/jes.202602105}
}
EUVIMED