Abstract
Early and efficient detection of gaps in doors is crucial for maintaining manufacturing plants, as it prevents energy loss, reduces equipment strain, and ensures optimal operating conditions, all while lowering maintenance costs and improving overall operational efficiency. This approach leverages advanced thermal imaging technology to identify air leaks and is more cost-effective than employing expensive ultrasonic devices and similar tools. Additionally, the model automates the detection process, eliminating the need for human intervention or proximity measurements, thereby enhancing efficiency and accuracy in identifying and addressing air leaks. The outcomes of this paper can serve as a foundation for advanced airflow quantification in structural gaps. The article focuses on developing a deep learning model to accurately identify and outline these gaps using a fusion of thermal and RGB (red, green, and blue) data. Understanding the relationship between physical gap dimensions and thermal characteristics is a critical step before estimating airflow rates in terms of cubic feet per minute (CFM). This initial phase shows promising results for future developments in estimating airflow, ultimately contributing to the optimization of energy usage and heating, ventilation, and air conditioning (HVAC) system regulation.