View in EDS

An hybrid fuzzy based deep convolutional neural networks for big data based hyper-spectral image classification in agricultural applications.

Saved in:
Bibliographic Details
Title: An hybrid fuzzy based deep convolutional neural networks for big data based hyper-spectral image classification in agricultural applications.
Authors: Barath, S.1 (AUTHOR) barath_s37@outlook.com, Ramesh, G.1 (AUTHOR)
Source: Australian Journal of Electrical & Electronic Engineering. Dec2025, Vol. 22 Issue 4, p628-640. 13p.
Subjects: Big data, Agricultural technology, Spectrum analysis, Quantitative research, Feature extraction, Image recognition (Computer vision), Fuzzy neural networks, Convolutional neural networks
Abstract: Hyperspectral image delves into hundreds of distinct spectral bands, unveiling an intricate 'fingerprint' of light reflected by crops, soil, and various environmental elements. This detailed view enables precise identification in the realm of agricultural practices. This information's intricate analysis and categorisation demand substantial computational prowess and storage capacity. The realm of big data frameworks and distributed computing realms has become paramount in efficiently managing this abundance. Within hyper-spectral images, pixels often encapsulate a medley of materials within a singular spatial entity, presenting a challenge in accurate pixel classification, particularly amidst intricate landscapes or overlapping spectral imprints. Furthermore, certain bands may prove superfluous or harbour irrelevant data. Tackling noise and redundancy necessitates employing refined preprocessing methods to enhance classification precision. To overcome the fallback in achieving improved accuracy, this proposed work uses a Hybrid Fuzzy Deep Convolutional Neural Networks (HFDCNN) model for classifying the big data of hyperspectral images. The proposed work is analysed using a real-time self-created dataset created using hyperspectral images of crops and soil across various parts of Tamilnadu. The proposed work is analysed in terms of accuracy, precision, recall, and F score, and it is compared with the existing state-of-the-art methods to prove its supremacy. [ABSTRACT FROM AUTHOR]
Copyright of Australian Journal of Electrical & Electronic Engineering is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Hyperspectral image delves into hundreds of distinct spectral bands, unveiling an intricate 'fingerprint' of light reflected by crops, soil, and various environmental elements. This detailed view enables precise identification in the realm of agricultural practices. This information's intricate analysis and categorisation demand substantial computational prowess and storage capacity. The realm of big data frameworks and distributed computing realms has become paramount in efficiently managing this abundance. Within hyper-spectral images, pixels often encapsulate a medley of materials within a singular spatial entity, presenting a challenge in accurate pixel classification, particularly amidst intricate landscapes or overlapping spectral imprints. Furthermore, certain bands may prove superfluous or harbour irrelevant data. Tackling noise and redundancy necessitates employing refined preprocessing methods to enhance classification precision. To overcome the fallback in achieving improved accuracy, this proposed work uses a Hybrid Fuzzy Deep Convolutional Neural Networks (HFDCNN) model for classifying the big data of hyperspectral images. The proposed work is analysed using a real-time self-created dataset created using hyperspectral images of crops and soil across various parts of Tamilnadu. The proposed work is analysed in terms of accuracy, precision, recall, and F score, and it is compared with the existing state-of-the-art methods to prove its supremacy. [ABSTRACT FROM AUTHOR]
ISSN:1448837X
DOI:10.1080/1448837X.2024.2430654