Research on Construction of Hyperspectral Sample Library for Insulators and Application of Lightweight Detection Device
Main Article Content
Abstract
This paper conducts research on the construction of a hyperspectral sample library for insulators and the development of a lightweight detection device. Firstly, the principle of hyperspectral imaging and the spectral response laws of different insulator defects are analyzed. A standardized experimental platform is built to complete the acquisition, preprocessing and annotation of multi-type, multi-defect and multi-working-condition data. The samples are expanded through spectral and spatial data augmentation, and a high-quality special hyperspectral sample library for insulators is constructed. Secondly, the detection model is lightweight optimized by combining pruning, quantization, knowledge distillation and other methods, which reduces the computational overhead of the model while ensuring the accuracy. Finally, the software and hardware co-design is completed, an integrated lightweight detection device is developed, and laboratory and field tests are carried out. The experimental results show that the sample library constructed in this paper has high annotation quality and good generalization performance. the inference speed of the optimized model meets the requirements of real-time detection. the self-developed device has low power consumption and strong portability, and has a high recognition rate for both macroscopic and hidden defects of insulators. Its comprehensive performance is superior to traditional detection methods, which can be effectively applied to the field inspection scenario of transmission lines and has high engineering application value.
Downloads
Article Details
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
References
E. Li, Q. Wang, R. Liu, et al., “A lightweight cross-scale reconstruction framework with synergistic attention for insulator and defect detection in power grid environments,” Eng. Appl. Artif. Intell., vol. 176, p. 114681, 2026, doi: 10.1016/J.ENGAPPAI.2026.114681.
M. Alanazi, “Adversarially robust and explainable insulator defect detection for smart grid infrastructure,” Energies, vol. 19, no. 4, p. 1013, 2026, doi: 10.3390/EN19041013.
Y. Fan, Y. Guo, Y. Liu, et al., “Study on the aging status of insulators based on hyperspectral imaging technology,” Opt. Express, vol. 32, no. 4, pp. 5072–5087, 2024, doi: 10.1364/OE.506030.
X. Du, X. Lei, T. Ye, et al., “A review of research on intelligent fault detection of power equipment based on infrared and voiceprint: methods, applications and challenges,” Glob. Energy Interconnect., vol. 8, no. 5, pp. 821–846, 2025, doi: 10.1016/J.GLOEI.2025.08.001.
G. B. Gang, “Advanced techniques in agricultural drone applications: Multispectral imaging, thermal analysis, precision landing, and customized pulley systems for enhanced remote sensing,” Int. J. Aeronaut. Space Sci., vol. 27, no. 1, pp. 1–14, 2025, doi: 10.1007/S42405-025-01005-6.
Q. Wang, Z. Fan, Z. Luan, et al., “Insulator abnormal condition detection from small data samples,” Sensors, vol. 23, no. 18, 2023, doi: 10.3390/S2 3187967.
G. Ran, W. Chenlin, L. Jinxiao, et al., “Lightweight fault diagnosis method in embedded system based on knowledge distillation,” J. Mech. Sci. Technol., vol. 37, no. 11, pp. 5649–5660, 2023, doi: 10.1007/S12206-023-1007-3.
V. Lodhi, D. Chakravarty, and P. Mitra, “Hyperspectral imaging system: Development aspects and recent trends,” Sens. Imag., vol. 20, no. 1, pp. 1–24, 2019, doi: 10.1007/s11220-019-0257-8.
U. I. M. Haque, R. Kannan, D. J. Hinkle, et al., “Predicting high-resolution spatial and spectral features in mass spectrometry imaging with machine learning and multimodal data fusion,” Adv. Intell. Discov., vol. 2, no. 2, p. e202500021, 2025, doi: 10.1002/AIDI.202500021.
J. Feng, J. Zhang, T. Li, et al., “Spectral–Spatial joint feature extraction for hyperspectral image based on high-reliable neighborhood structure,” IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens., vol. 14, pp. 9609– 9623, 2021, doi: 10.1109/jstars.2021.3112158.
Y. Ou, Y. Guo, Y. Fan, et al., “Non-destructive testing of defects in composite insulators: A systematic review,” Measurement, vol. 272, p. 121053, 2026, doi: 10.1016/J.MEASUREMENT.2026.121053.
P. Yu, H. Wang, J. Han, et al., “Research on typical defect identification technology of composite insulators for ultra high voltage transmission lines based on spectral feature extraction,” Appl. Math. Nonlinear Sci., vol. 10, no. 1, p. 20250956, 2025, doi: 10.2478/AMNS-2025-0956.
D. Cozzolino, P. J. Williams, and L. C. Hoffman, “An overview of pre-processing methods available for hyperspectral imaging applications,” Microchem. J., vol. 193, 2023, doi: 10.1016/J.MICROC.2023.109129.
N. T. Karadeniz, S. Ekici, and E. Avci, “Deep learning-based detection of power transmission lines using YOLOv4 and YOLOv8,” Sci. Rep., vol. 16, no. 1, p. 2382, 2025, doi: 10.1038/S41598-025-32200-W.
C. Lanhang, “Defect identification of electricity transmission line insulators based on the improved lightweight network model with computer vision assistance,” Heliyon, vol. 10, no. 10, p. e30405, 2024, doi: 10.1016/J.HELIYON.2024.E30405.