EUVIMEDEUVIMED entdecken
Crossref

A Novel CNN-Based Dual-Stream Network with Deeplabv3+ for Accurate Fire and Animal Detection in Foggy Environments

Shafiya Shaik, K. Sujatha

Journal of Circuits, Systems and Computers · 25.04.2026

DOI: 10.1142/s0218126626501690 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Circuits, Systems and Computers, online erschienen am 25.04.2026. Verfasst von Shafiya Shaik und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 43 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Fire and animal detection under extremely aggravated atmospheric fog conditions is very challenging due to low visibility, noise and environmental hindrances. These create significant obstacles to the reliability and accuracy of detection systems, which are vital for wildlife conservation and disaster management. This study proposes a dual-stream network with DeepLabv3[Formula: see text] to enhance certainty and robustness under poor foggy conditions. The framework includes a multi-scale fusion dehazing network (MSFDNet) for dehazing, nonlocal means (NLM) for denoising and data augmentation techniques. DeepLabv3[Formula: see text] with enhanced atrous spatial pyramid pooling (ASPP) ensures precise segmentation. Feature extraction uses spider wasp optimization (SWO) and the hippopotamus optimization algorithm (HOA). Detection integrates YOLOv8-SE, Enhanced FocalNet and convolutional neural networks-long short-term memory-multi-head self-attention (CNN-LSTM-MSA) for reliable classification performance. The proposed system achieves a notable improvement, reaching 99.2% accuracy, demonstrating its effectiveness for detection under severe fog conditions.

Zitieren

APA

Shaik, S., & Sujatha, K. (2026). A Novel CNN-Based Dual-Stream Network with Deeplabv3+ for Accurate Fire and Animal Detection in Foggy Environments. Journal of Circuits, Systems and Computers. https://doi.org/10.1142/s0218126626501690

Vancouver

Shaik S, Sujatha K. A Novel CNN-Based Dual-Stream Network with Deeplabv3+ for Accurate Fire and Animal Detection in Foggy Environments. Journal of Circuits, Systems and Computers. 2026. doi:10.1142/s0218126626501690

BibTeX
@article{Shaik2026,
  author = {Shaik, Shafiya and Sujatha, K.},
  title = {A Novel CNN-Based Dual-Stream Network with Deeplabv3+ for Accurate Fire and Animal Detection in Foggy Environments},
  journal = {Journal of Circuits, Systems and Computers},
  year = {2026},
  doi = {10.1142/s0218126626501690}
}
Kostenlos starten →