Research on Target Tracking Method Based on Infrared Video from Photoelectric Pod

The infrared target tracking of airborne optoelectronic pods faces the challenges of weak targets, background interference, scale changes and target occlusion, etc. In order to meet the real-time and high-precision target tracking requirements for optoelectronic pods on the future battlefield, combined with the characteristics of the existing equipment capabilities, this paper proposes an infrared target identification and tracking algorithm based on static invariance of the silhouette and dynamic continuity. The algorithm constructs a target contour model and motion characteristic model within the infrared video signal, and designs relevant decision-making criteria to achieve fast, accurate, and stable recognition and tracking of small targets. Through continuous tracking experiments of airborne targets and ground targets, compared with the traditional feature point matching and deep learning-based ECO algorithm, this algorithm shows significant advantages in accuracy and timeliness, and is able to stably track the variable-scale targets in complex scenes, which provides a new way of thinking and methodology for the target identification and tracking of airborne targets, and is of great significance for improving the performance of the optoelectronic pods and expanding their application fields. It is of great significance to improve the performance of optoelectronic pods and expand their application fields.