Research On Detection And Tracking Algorithm Of Laser Strike On Small Targets

The "low,small and slow" target has slow flight speed,small scattering area and low flight altitude.At the same time,it also has strong mobility,which poses a great threat to people’s lives and property.Therefore,many researches on laser strike system have been carried out for "low,small and slow" flying objects at home and abroad.Some research results can identify and strike "low,small and slow" targets in time to varying degrees.As a part with the function of "eye" in the strike system,infrared imaging system has become one of the main research contents in the strike system because of its strong concealment and adaptability to small target detection.It is very important for the whole laser strike system to detect and track "low,small and slow" and other small targets in time.The "low,small and slow" and other small target occupies less pixels in the imaging system and is in a complex background.The target is easy-to be confused with the influencing factors,which will cause false alarm in the process of detection and tracking.Therefore,it is necessary to customize a set of detection and tracking algorithms according to the characteristics of "low,small and slow" and other small targets.In this paper,the infrared small target detection and tracking technology is studied.In view of the difficulties in traditional infrared small target detection,which are difficult to improve the significance of the target and extract the target characteristics,slow processing speed and difficult to meet the difficulties of accurately identifying and tracking the target under complex background,a set of targeted infrared small target characteristics detection and tracking method is proposed.Firstly,the potential target area is obvious through fractal filtering,the cross median filter is used to remove the high-frequency noise,,the image is segmented by morphological gradient,and the feature vector including the centroid of the candidate target area is constructed as the feature of the target area.The distance between the target areas is calculated according to the feature vector,and the candidate areas are merged by setting an adaptive threshold,Finally,SVM,a pattern recognition method,is used to classify the target.Compared with some traditional infrared small target detection methods,its signal-to-noise ratio gain and background suppression factor are improved,and the detection speed is significantly improved.In terms of tracking,because each feature of the target has been analyzed in detail,the method of constructing target chain is used to track the target.The dynamic target recognition method can transfer the target feature information through the construction of target chain and track the target in video.Through the research and design of infrared small target detection and tracking system,the detection and tracking module is designed in detail.The target detection and tracking unit is mainly composed of infrared camera,and the chip selection is explained from the aspects of image preprocessing and target recognition.At the same time,the system experimental platform is built,and the simulated spot experiment is carried out to verify it.The experimental results show that the system can capture the simulated small target under the complex background and track the target in real time when moving.At the same time,it has high resolution,the frame rate can be maintained at 25 to 30 frames,the accuracy can reach 1 pixel,the detection rate can reach 98%,and the system is stable.It can meet the needs of the actual situation.