The Research On Infrared Target Detection And Tracking Technology

With the development of infrared technology,infrared small target detection and tracking has been widely used in medicine,military and other fields.During the infrared imaging process,when the target is far from the infrared detector,as interfered by the noise of infrared detector and atmospheric heat radiation,the infrared target image is small with blurred target boundary,low contrast,low signal-to-clutter ratio and lack of efficient information such as color and texture,which constitutes great challenges to the detection and tracking of infrared small targets.Therefore,the research on detection and tracking of infrared small target is of great significance.This thesis has deeply studied the infrared image preprocessing,the infrared small target detection and tracking.In the infrared image pre-processing stage,the classical top-hat morphological operator is used to preprocess the original image and suppress the relevant background noise.In the infrared small target detection stage,based on the principle of human visual contrast mechanism,the contrast characteristic of the infrared target is used to calculate the saliency map of the image,after the target is enhanced and the background is suppressed,the ROI of the image is extracted to determine the candidate infrared small target,and then by extracting the contour of the candidate target to adaptively construct the dual-type morphological structure operator.In the region of interest(ROI)within the local range,through suppressing clutter and noise by the morphological filtering of the adaptive dual-type structure element and the median filter algorithm,the small target detection is achieved.In the infrared small target tracking stage,the classical Spatio-Temporal Context(STC)tracking algorithm is improved,Based on the original algorithm with the gray feature,the local weighted gray level information entropy(LWIE)which can represent the degree of gray value change in the local region is added,And with the local contrast characteristics of the target and thesurrounding background,an evaluation model for each target of the current tracking is established,thus determining whether the currently tracked target is accurate or effective.In this thesis,the proposed algorithm will conduct comparison test for single-objective and multi-objective under the complex background such as in the sea level,sky and so on,and then calculate the detection rate,false alarm rate and other performance indicators.The experimental results show that the proposed algorithm can adapt well to different complex background conditions,and has good validity,timeliness and robustness.