MineSeeker: A Novel Design for Detecting Explosive Mines Using Low-Resolution Thermal Imaging

Landmines, particularly the PFM-1 Butterfly landmine, continue to pose a significant threat to civilian populations in conflict zones around the world. Traditional detection methods, such as manual searches, drone surveillance, and optical detection technologies, are often rendered ineffective due to the PFM-1's camouflaged coating and small size. In response to these challenges, our research introduces a novel approach utilizing an enhanced version of the YOLOv8 deep learning algorithm, integrated with cubic spline interpolation, to improve detection accuracy with low-resolution thermal imagery. Our innovative algorithmic approach enables the development of a cost-effective and compact mine detection system. This system employs affordable, low-resolution thermal cameras and is designed to operate efficiently on low-computational power devices like the Raspberry Pi. The integration of cubic spline interpolation enhances image resolution, allowing for more accurate detection of mines despite the limitations of low-resolution thermal imagery. Experimental results have demonstrated the effectiveness of our proposed design, showcasing its ability to offer a scalable and efficient solution for mine detection. The enhanced YOLOv8 algorithm significantly improves detection accuracy, making the system viable for deployment in various operational environments. Additionally, the adaptability of this approach extends its potential applications beyond mine detection, suggesting its utility in other object detection scenarios. By providing a more accessible and reliable method for detecting landmines, our system holds the promise of reducing the risk to civilian populations in affected areas. The advancements presented in this research not only address the immediate need for improved mine detection but also contribute to the broader field of object detection technology, potentially influencing future developments in related areas.