AL-Rafidain Journal of Computer Sciences and Mathematics

Cigarette smoking poses a major challenge to public health worldwide and has serious health consequences. Recent advances in deep learning, machine learning, Artificial Intelligence (AI), big data analytics, and computer vision have greatly enhanced smoking detection. These technologies enable the analysis of diverse datasets to identify patterns that indicate smoking behavior By enhancing the effectiveness of smart smoking detection systems and so we can better protect public health and reduce exposure to secondhand smoke in public places. An AI-based monitoring system is crucial to enhancing the fight against smoking in restricted areas by establishing a framework for identifying the locations of smoke detection systems across the city. This paper aims to shed light on the effectiveness of these smart systems in facilitating smoking cessation efforts and ensuring compliance with no smoking rules by reviewing previous studies on smoker detection and the algorithms used in those studies and their degree of effectiveness and efficiency in achieving the intended goal. They were examined, analyzed, classified, A comparison was made between research that used machine learning and research that used deep learning, and a comprehensive scientific comparison was made, with special attention paid to the data used to build the model. Furthermore, this paper will provide data on the results of indoor and outdoor smoker detection using smart algorithms, contributing valuable insights for future research in this area.