DEEP Learning-based air quality monitoring model via BM-KMC using seasonal images

R. Udaya Shanmuga () and G. Tamilpavai ()
Additional contact information
R. Udaya Shanmuga: Anna University
G. Tamilpavai: Anna University

Environment, Development and Sustainability: A Multidisciplinary Approach to the Theory and Practice of Sustainable Development, 2025, vol. 27, issue 9, No 103, 23001 pages

Abstract: Abstract Nowadays, numerous cities have seen severe air pollution (AP) that affects the daily lives of people and produces harmful health impacts. As a result, researchers have shown considerable interest in monitoring AP. However, accurate monitoring of air quality remains a significant challenge. To overcome this problem, an image-based analysis framework is proposed here to monitor air quality. For this purpose, initially, input images relevant to diverse seasons are collected from publicly available sources. Then, the input images are converted into a hue, saturation, and intensity (HSI) color transformation model for easily analyzing and processing the colors independently. Afterward, the HSI images are enhanced by employing the HMFBHE algorithm for extracting information effectively in night scene images. Next, the sky segmentation is performed on the contrast-enhanced images by using the gibbs generalized gradient vector flow (GGGVF) algorithm, which is done by means of pixel values. Then, to evaluate the linear relationship’s strength betwixt the variables, the correlation calculation is performed on the segmented sky images. Afterward, the correlated images are clustered as strong, medium, and low by employing the broyden model-based K-means clustering (BM-KMC). After that, the low correlation valued features are extracted from the clustering phase. In this way, from the other objects of the segmented sky image, the features, including buildings, trees, and so on are extracted. Thereafter, the optimal features are chosen from the extracted features by employing the ratio nambie beetle optimization (RNBO) technique. At last, from the selected features, the air quality is monitored as good, satisfactory, moderate, poor, very poor, and severe by using the LPKF-PSELU-ResNet algorithm. Thus, the experimental outcomes proved that the proposed model attained a high accuracy of 0.98677, which outperformed the prevailing techniques.

Keywords: Air pollution (AP); Broyden model-based K-means clustering (BM-KMC); Deep learning (DL); Lecun polynomial kernelized function-perturbation scaled exponential linear unit-residual network (LPKF-PSELU-ResNet); Preprocessing; Gibbs generalized gradient vector flow (GGGVF); Hessian matrix form of bi-histogram equalization (HMFBHE) (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10668-024-05745-5 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:endesu:v:27:y:2025:i:9:d:10.1007_s10668-024-05745-5

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/10668

DOI: 10.1007/s10668-024-05745-5

Access Statistics for this article

Environment, Development and Sustainability: A Multidisciplinary Approach to the Theory and Practice of Sustainable Development is currently edited by Luc Hens

More articles in Environment, Development and Sustainability: A Multidisciplinary Approach to the Theory and Practice of Sustainable Development from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().