With the increasing availability of high-dimensional hyperspectral data from modern remote sensing platforms, accurate and efficient classification methods are urgently needed to overcome challenges such as spectral redundancy, spatial variability, and the curse of dimensionality. The current hyperspectral image classification technique has become a crucial tool for analyzing material information in images. However, traditional classification methods face limitations when dealing with multidimensional data. To address these challenges and optimize hyperspectral image classification algorithms, this study employs a novel fusion method that combines principal component analysis (PCA) based on null spectral information and 2D convolutional neural networks (CNNs). First, the original spectral data are downscaled using PCA to reduce redundant information and extract essential features. Next, 2D CNNs are applied to further extract spatial features and perform feature fusion. The powerful adaptive learning capabilities of CNNs enable effective classification of hyperspectral images by jointly processing spatial and spectral features. The findings reveal that the proposed algorithm achieved classification accuracies of 98.98% and 97.94% on the Pavia and Indian Pines datasets, respectively. Compared to traditional methods, such as support vector machines (SVMs) and extreme learning machines (ELMs), the proposed algorithm achieved competitive performance with 98.81% and 98.64% accuracy on the same datasets, respectively. This approach not only enhances the accuracy and efficiency of the hyperspectral image classification but also provides a promising solution for remote sensing data processing and analysis.
Distributed Acoustic Sensing (DAS) systems face increasing challenges in massive data processing and real-time fault diagnosis due to the growing complexity of industrial environments and data volume. To address these issues, an end-to-end diagnostic framework is developed, integrating Mel-Frequency Cepstral Coefficients (MFCCs) for high-efficiency signal compression and Liquid Neural Networks (LNNs) for lightweight, real-time classification. The MFCC algorithm, originally used in speech processing, is adapted to extract key features from DAS vibration signals, achieving compression ratios of 60–100× without significant information loss. LNNs’ dynamic topology and sparse activation enable high accuracy with extremely low latency and minimal computational cost, making it highly suitable for edge deployment. The proposed framework was validated both in simulated environments and on a real-world conveyor belt system at Qinhuangdao Port, where it achieved 100% accuracy across four vibration modes over 14 weeks of operation. Comparative experiments show that LNNs outperform traditional models such as 1D-CNN and LSTMs in terms of accuracy, inference speed, and model size. The proposed MFCC-LNN pipeline also demonstrates strong cross-domain generalization capabilities in pipeline monitoring, seismic detection, and speech signal processing.