Enhancing Thermal Object Detection in Adverse Weather Conditions for Automotive Applications Using Deep Learning-Based CNN and SVM Models

Northeastern Indian cities Such as Delhi have harsh weather conditions, and high air pollution often results in poor visibility, leading to numerous accidents. This paper proposes a thermal object detection system that leverages thermal imaging and deep learning algorithms to detect vehicles and objects based on their heat signatures to address this issue. Unlike traditional vision-based systems that rely on visible light, this approach is resilient to low-light conditions, shadows, and glare, ensuring reliable detection even in challenging environments. The system is tested on real-world images, demonstrating its effectiveness in traffic management, security, and industrial automation. Additionally, thermal imaging technology provides advantages such as detecting objects at longer ranges and offering insights into temperature and material properties. These findings show the potential to significantly enhance safety and security in scenarios where visibility is compromised.