<![CDATA[Smoke-Free Areas are implemented to protect public health; however, monitoring and evaluating compliance remain challenging due to the lack of automated and spatially integrated monitoring systems. This study aims to develop a mapping and classification system for Smoke-Free Area (KTR) compliance using the YOLOv11 object detection algorithm and Geographic Information System (GIS)-based spatial analysis on environmental images. Data collection was conducted in Banda Sakti District, Lhokseumawe, at nine observation locations consisting of places of worship, public open spaces, and workplaces. The dataset consisted of three object classes, namely smoking activity, cigarette, and ashtray, combined with spatial variables such as latitude and longitude, where latitude represents the north–south geographic position and longitude represents the east–west geographic position of each observation point. These spatial variables were integrated with the YOLOv11 detection results to enable the mapping and visualization of KTR violations within the GIS environment. The YOLOv11 model was evaluated using precision, recall, mAP50, and mAP50-95 metrics. Experimental results showed that the model achieved a precision value of 0.772, recall of 0.741, mAP50 of 0.770, and mAP50-95 of 0.602, indicating moderate object detection performance under various environmental conditions. Spatial analysis results revealed that out of 145 observation points, 112 points were categorized as major violations, 25 points as minor violations, and only 8 points as compliant areas. Therefore, the integration of YOLOv11 and GIS provides a digital-based approach for supporting Smoke-Free Area compliance monitoring and spatial analysis.]]>
Copyrights © 2026
