Development Of Infrared Small Target Tracker System

The detection and tracking of infrared small targets is a core technology in the ground-to-air early warning system.In the early warning process,the farther the target is from the infrared sensor,the fewer pixels are occupied in the infrared image and the lower the contrast is.Therefore,in order to detect the enemy target as soon as possible,the infrared small target detection and tracking algorithm need to be studied.For this reason,this project developed a small infrared target tracking system based on DSP to develop a small infrared target tracking system.Firstly,the demand analysis of the infrared small target tracking system is carried out,and the technical indicators of the infrared small target tracking system are given.Aiming at the requirements of the requirements and the system performance indicators for the hardware platform power consumption and processing speed,the appropriate hardware platform was selected,and the overall scheme of the infrared small target tracking system was designed.Research shows that this design can solve the problem of mismatch between algorithm and resources.Secondly,the system resources of the hardware platform are comprehensively analyzed.The definition of infrared small target is given,and the infrared small target detection and tracking algorithm are analyzed in detail.The appropriate small target detection algorithm is selected.For traditional threshold segmentation,the segmentation effect of pre-processed small-target images is poor.In this paper,an adaptive threshold segmentation method based on average maximum value for infrared small-object segmentation is designed.And based on the analysis of the tracking algorithm,a suitable motion model was established.After determining the small target detection and tracking algorithm,the complexity of the algorithm is analyzed to ensure that it can be implemented on the hardware platform.And for the problem that cloud cover obstructs the loss of target,the algorithm is optimized and a loss reacquisition method is designed.The experimental results show that the algorithm can improve the tracking accuracy under the condition that the target is occluded.Again,a small target tracking system was implemented on the hardware platform.Firstly,the image transmission interface based on Gigabit Ethernet interface is designed,and a multi-core cooperative mechanism based on shared memory is designed to make reasonable use of the computing resources of multi-core DSP.And using a compiler to optimize the entire system.Through testing and analysis,the system can meet the requirements of processing speed andtracking accuracy in technical indicators.Finally,according to the requirements of system speed and reliability in the technical indicators of the system,the entire system was tested for functionality and performance.The system functions and system performance under the lab environment and the field environment were tested separately.The test results are analyzed in detail.The test results show that the infrared small target system developed by this project has good robustness and adaptability.After obstructing the target clouds,the target can be lost and the target can be recaptured.,And all functions can be achieved,tracking speed and tracking accuracy can meet the requirements of technical indicators.