Research On Algorithm Of Detecting Infrared Moving Target In Low Altitude Background

Infrared moving target detection in a complex background is one of the hot topics in the field of modern image processing.Long distance flying targets in sea-sky background with few pixels in infrared image,serious energy,lack of geometry and information of structure and texture have less information for detection systems to use.And the complex interference from natural and artificial leading to low SNR(signal-to-noise ratio)greatly affect the target detection.It is of great significance to explore and study new small target detection algorithms and how to apply existing processing methods to small target detection.This thesis made a comparative study of some target detection methods propounded by domestic and international researchers.Based on the infrared imaging search and tracking device,which being the project background,this thesis designed a set of detection algorithm of dim flying target under sky-ground background including image pre-processing,single frame detection and multi-frame confirmation,etc.At first,the characteristics of typical flying targets under low-altitude background and the effects from natural and artificial interference are analyzed in detail.For these characteristics,an improved morphological filtering method based on background estimation is proposed.In the method of extracting the ground-sky line,several common methods of extracting straight lines in image processing is implemented,and then the scene characteristics in the image to be processed and the advantages and disadvantages of several methods are integrated,and the original method is improved.The improved line mean gradient method and the straight line fitting method are combined.Improve the accuracy of ground-sky line detection.Through the performance comparison between several segmentation methods,an improved iterative threshold segmentation is selected to detect small single-frame targets with high probability after extracting the ground-sky line;finally,this thesis combines the multi-feature correlation method into traditional pipeline filtering.Successfully confirm the goal.Simulations were introduced in order to validate the above-mentioned the methods and the whole algorithm.The algorithm designed in this paper can meet the requirements of the device platform,immediacy performance requires the image processing frequency to reach 100 Hz,and the target detection probability to reach 90%.The analysis of the experimental results shows that,compared with the algorithm on the original platform,the calculation time is reduced by 0.002 s,and the target detection probability is increased by 5%.The target detection can be achieved under the condition of extremely low signal-to-noise ratio of small flying targets.The design of this paper is based on the design of an existing vehicle-mounted infrared search and tracking device.Based on the algorithm of this device,other algorithms are tried to improve the target detection performance of the device and achieve the improvement of the target detection performance in the algorithm.