Research On Detection Method Of Infrared Moving Small Target In Complex Scene

Infrared small target detection has been widely used in remote monitoring and military reconnaissance fields.In these practical applications,due to the long imaging distance,the size of the acquired target is very small,and there is no prior knowledge of the original shape and movement state.In addition,the actual imaging scene is complex and changeable,which makes the signal-to-clutter ratio of the small target very low and is easily submerged by background clutter.Traditional small target detection methods cannot effectively detect small infrared moving targets in complex backgrounds.Therefore,how to detect small infrared moving targets in complex backgrounds with high reliability is considered to be a difficult task.Starting from the analysis of infrared image characteristics,this dissertation conducts in-depth research and analysis on image background suppression,adaptive parameter selection and small target detection,building a set of detection methods for moving small targets under complex backgrounds.The main research results are summarized as follows:1.In order to make full use of the spatiotemporal information in the image sequence,and effectively suppress the image background obtained from the complex scene and enhance the small target signal,a method of infrared image background suppression based on spatiotemporal three-dimensional information extraction is designed.Firstly,the Fast Fourier Transform is utilized to process the infrared image;then,according to the corresponding characteristics of different components in the image,the spatiotemporal three-dimensional information extraction method is designed to suppress the complex background of the image;finally,according to the statistical characteristics of small targets and noise in the infrared image,a method to realize image difference and stack is designed to enhance the small target signal.The experimental results show that the proposed method can not only effectively suppress the background and noise of the image,but also enhance the signal of the small target.2.In order to obtain the parameters of infrared image background suppression method adaptively according to image information,an adaptive infrared image background suppression method based on improved particle swarm optimization algorithm is designed.This method transforms the adaptive parameter selection problem into a multi-objective optimization problem.In the multi-objective optimization problem,this method first uses a local optimization strategy based on the human visual system to obtain coordinate points suitable for optimization;then,an improved adaptive environment selection particle swarm optimization algorithm is designed to obtain the Pareto front of the optimization objectives;finally,an improved knee point selection strategy is utilized to select the most suitable knee point from the front and obtain its corresponding background suppression method parameters.Comprehensive experimental results show that the adaptive infrared image background suppression method has better background suppression performance than the baseline method and the background suppression method based on empirical acquisition method parameters.3.In order to improve the accuracy of small moving target detection and eliminate residual noise,based on the designed image background suppression method,a small target detection method based on inter-frame correlation and energy accumulation is designed.This method first determines the suspicious motion trajectory of the small target according to the inter-frame information of the small target;then,based on the acquisition of the motion trajectory,a method based on energy accumulation is designed to enhance the small target signal;finally,the small target detection is realized according to the quantitative analysis of the suspicious trajectory and the constant false alarm rate judgment method.Comprehensive experimental results of real scenes show that this method can stably detect moving small targets in different complex scenes.At the same time,the proposed method has the characteristics of high detection rate and low false alarm rate.