FPGA Implementation of Infrared Images Small Targets Track-Before-Detect System

Nowadays, image processing and computer vision play important roles in many applications, such as early warning, precision guidance, and safety precautions, etc. Also, imaging equipment is often used as the assistance for these applications since infrared technology can be used for various weather conditions. Hence, the application of infrared images for target detection and tracking is an important part. Our proposed method adopts Laplace of Gaussian filter and template matching to effectively detect small and nearly invisible targets. Additionally, the application of Gaussian window overcomes the problem of dim targets. Similarly, using image sequences to detect moving targets may avoid the false alarms due to complicated background. In addition, the infrared small target detection system needs to reflect the detection in time so that real-time effect is indispensable. We use the high-level synthesis (HLS) tool to implement the proposed method and verify on the FPGA experimental platform. Finally, the experimental results show that the proposed infrared small target detection system in this research can effectively detect small objects and demonstrate on the FPGA platform in real time.