Evaluation of a target detection algorithm using the infrared sensor model in real time

In this work, the performance of a target detection algorithm to be used against maritime platforms is evaluated in real-time. The software developed in this work includes an infrared (IR) Scene Generator Software, a Heat Transfer Software, an IR Sensor Model Software and a Target Detection Software. In order to enable all software components to operate in real time for large format images, the routines that can be operated in a parallel fashion are computed in the Graphics Processing Unit (GPU). Image degradations (blur and noise) with physics-based models are applied to dynamic images of the target generated by the IR Scene Generator and their effect on the performance of the target detection algorithm has been evaluated. We believe that the results of this work will shed light to the efforts for increasing the efficiency of target detection, recognition, and tracking algorithms and also to the efforts for the development of the hardware-in-the-loop systems for dynamic performance testing which have a need to operate in real time.