Evaluation of insulator aging status based on multispectral imaging optimized by hyperspectral analysis

Tracking moisture contents in the pollution layer on a composite insulator surface using hyperspectral imaging technology

“…Characteristic band selection, which aims to reconstruct the original information from low-dimensional data by capturing the most informative features, is considered an effective method for reducing acquisition time and hardware requirements while minimizing computational resources. [29][30][31] In this paper, Successive Projections Algorithm (SPA), Competitive Adaptive Reweighted Sampling (CARS) and Random Frog (RF) were adopted to select characteristic bands, respectively. Similarly, PLSR models were adopted to evaluate the dimensionality reduction results.…”

“…Selection of the characteristic bands is crucial for minimizing redundant information in hyperspectral images and optimizing the cost-effectiveness of imaging equipment, both in terms of software and hardware. 31 This study employed SPA, CARS, and RF methods to individually identify characteristic bands from the full spectra. Fig.…”

Visual characterization of pollution and moisture parameters of the polluted HTV silicone rubber surface with hyperspectral imaging technology

“…Recently, scholars have preliminarily explored ESDD and MC pattern recognition and visualization methods by incorporating the spectral database and prior knowledge, inspired by the application of hyperspectral imaging (HSI) technology for fine nondestructive testing [11][12][13][14]. Qiu et al [7] first combined HSI and extreme-learning-machine-based classification method for ESDD detection on the insulators surface, achieving an impressive classification accuracy of over 87.5%.…”