Research On Target Detection Algorithm In Infrared Warning System

As an important part of opto-electrical countermeasure technology, infrared warning system plays an invaluable role in modern warfare. Infrared target detection is an important technology in infrared warning system, and the performance of its algorithm directly determines the warning distance and the detection probability of infrared warning system. Therefore, infrared target detection algorithm occupies a pivotal position in infrared warning system.The quality of infrared images directly impacts on the performance of the follow-up target detection algorithm. So the paper expound the image quality evaluation indices firstly, then infrared image quality analysis software is designed and implemented. On this basis, the dead-pixel detection and compensation algorithm in the image preprocessing is purposefully researched.In order to solve the problem that it's hard to define threshold in threshold segmentation in Detect Before Track (DBT), a DBT algorithm based on Particle Swarm Optimization (PSO) is put forward. This algorithm adopts fixed weights filter to predict the background of image firstly, and then finds the optimal threshold of the differential image between the raw image and the background image by PSO and does do threshold segmentation for differential image to get target location. What is more, with the simulation experiment of the algorithm, it is proved to be effective by analyzing the experimental results quantitatively and qualitatively.To ensure the system real time and detection probability, an infrared multiple-target real-time detection algorithm is studied, the algorithm adopts gradients filtering method to do image preprocessing and then do threshold segmentation for result image to get location of candidate targets. Upon that, data association is done for candidate targets in image sequences to determine the real targets. This algorithm is implemented in hardware system based on DSP and proved to be effective and superior through qualitative and quantitative analysis of experimental results.