Development Of Line Array Scanning Air-to-air Infrared Small Target Capture Module

The air-to-air infrared small target acquisition technology is widely used in both military and civilian fields,and it is one of the important monitoring methods for protecting the safety of airspace environment.Since the infrared small target reflects the characteristics of less pixels and low contrast in the imaging,it is self-evident that the real-time image processing system realizes the difficulty of capturing it.Aiming at the problem of capturing non-cooperative mobile infrared small targets in the vehicle platform,the overall scheme of FPGA+dual-DSP processing architecture was designed from the perspective of heterogeneous computing coordination.The overall framework of 8-core DSP collaborative processing architecture was designed from the perspective of task parallelization scheduling.In this paper,a small infrared target acquisition algorithm is designed from the perspective of algorithm implementation.Aiming at the realization of the target capture task collaboration technology,firstly,for the system computing resource limitation problem,the FPGA+dual-DSP heterogeneous computing task collaborative processing architecture is proposed.The FPGA as the main control unit controls the data transmission,and the DSP as the slave unit performs image processing and target acquisition.After a long-term loopback test in the field test,this heterogeneous computing collaborative architecture can complete processing tasks with limited computing resources.Secondly,aiming at the real-time data processing requirements of the system,a task parallelization scheduling collaborative processing architecture based on 8-core DSP is proposed.The architecture adopts a master-slave parallel programming model.The master core controls the data interaction and starts from the core,and the core responds to the master core control commands and completes the corresponding processing operations.After offline testing and field test online test,the collaborative architecture can meet the requirements of real-time processing of data in the processing of wide field of view and narrow field of view.Finally,based on the problem of switching between two kinds of fields of view,the system has designed different processing architectures for wide and narrow fields of view.For the wide field of view with low frame rate and large data volume,the principle of equalization is adopted.For the narrow field of view with high frame rate and small data volume,the principle of "ping-pong operation" is adopted.Through offline testing and field test online test verification,this processing architecture meets the requirements of system wide and narrow field of view on-demand switching.Aiming at the requirements of system generalization under complex conditions,an algorithm optimization for small-air infrared target acquisition under complex conditions is proposed.This topic considers two kinds of complex situations,including detectors and environmental disturbances and non-cooperative target mobility changes.After analyzing the causes,the corresponding treatment methods are proposed.After offline testing and field test online test,the wide field of view and narrow field of view meet the system requirements in terms of capture rate.Finally,the applicability of the system is evaluated.From the aspects of basic function,robustness,real-time and capture rate,the integrated field test of the air-infrared small target capture module designed by this subject is carried out.After all-round test and verification,this book The linear array scanning-to-air infrared small-target acquisition module designed by the subject satisfies the requirements of functions and performance,and has the conditions for practical application of the project.