Citation

Ahmad Kamsani, Hamzah Asadullah and Lye Abdullah, Mohd Haris (2025) Real-Time Cigarette Smoking Detection On Raspberry Pi Using YOLOv11. In: 2025 Multimedia University Engineering Conference (MECON), 21-23 July 2025, Cyberjaya, Malaysia.

Text 463.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Smoking poses significant health and environmental risks to individuals, bystanders, and the wider community. To mitigate smoking in public spaces, real-time monitoring systems can serve as effective preventive tools. With recent advancements in computer vision and deep learning, object detection from images and video streams has become increasingly accurate and efficient. This paper presents a real-time cigarette detection system using the YOLOv11 model, a single-stage object detection algorithm, optimized for detecting small objects. The trained model is deployed on Raspberry Pi for efficient edge inference. The proposed system achieves a mean Average Precision (mAP@50) of 0.892 and a detection speed of 2.3 frames per second (FPS) in video streams when converted to the NCNN deployment format. These results highlight the feasibility of deploying lightweight deep learning models on embedded systems for effective cigarette detection in various public environments.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	cigarette detection, real-time object detection, YOLOv11, computer vision, deep learning
Subjects:	Q Science > Q Science (General)
Divisions:	Faculty of Artificial Intelligence & Engineering (FAIE)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	18 Mar 2026 00:09
Last Modified:	18 Mar 2026 00:09
URII:	http://shdl.mmu.edu.my/id/eprint/15507

Downloads

Edit (login required)