Font Size: a A A

Research On Infrared Dim And Small Target Detection In Complex Background

Posted on:2016-08-07Degree:MasterType:Thesis
Country:ChinaCandidate:M D SiFull Text:PDF
GTID:2348330536967498Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
Dim and small target detection technology in complex background is key to information processing in IRSS,it's important to improve the system's performance.Because of very far away between the sensor and plane,the target on the plane like a point or spot lack some feature like shape and texture.It's very difficult to accomplish detect dim and small target in complex background.In this thesis,the infrared dim and small target detection algorithm combine with spatial and temporal to be studied.The main contents in this paper composed of three parts:In the second chapter,analyzes the infrared imaging model and the characteristic of target,noise and background based on the law of the infrared radiation.Then,introduces spatial,transformation and temporal domain infrared image pre-processing algorithm respectively.Lastly,discusses two mainly category target detection algorithms deeply in order to provide theory of foundations to infrared dim and small target detection.In the third chapter,it has studied the dim target detection algorithm based on visual saliency and image local contrast of a single image,proposes a improve frequency residual to suppress the background clutter in infrared image.Then,extract ROI and use image local contrast to decrease the false alarm.Analysis of the experiment results show proposed algorithm improve the performance compare to other common algorithms.In the forth chapter,a combine with spatial and temporal pre-processing method is proposed.Firstly,analyzes the characteristics temporal profile in different region in the image,proved the possibility to detect the dim and small target in temporal domain.Secondly,temporal variances of the blocks surrounding the window are calculated in temporal profile.Then,combine with temporal and spatial and get the fuse image.Analysis of the experiment results show proposed algorithm improve the performance to detect the target.Algorithms for target detection in this thesis can be used for infrared image sequence,and it provides support for the IRSS system.
Keywords/Search Tags:Infrared image, Dim small target, Visual Saliency, Temporal Profile, Spatial and Temporal
PDF Full Text Request
Related items