本文選題：SAR圖像 + 目標檢測　； 參考：《電子科技大學》2015年碩士論文

【摘要】：合成孔徑雷達(Synthetic Aperture Radar,SAR)的工作不受時間和天氣的影響,并且具有很強的透射性,憑借這些優(yōu)勢,它在軍事領域占據(jù)重要地位,軍事作戰(zhàn)提供軍事作戰(zhàn)所需要的偵察和目標探測信息,同時也普遍應用于氣象預告、地形地貌研究、道路檢測、災情監(jiān)控等民用方面。SAR目標檢測是利用目標的灰度、紋理、形狀、邊緣和方向等信息在SAR圖像中確定其位置,將目標與背景分離的圖像處理技術。隨著SAR圖像分辨率不斷提高,目標信息呈現(xiàn)爆炸性增長,如何快速有效地對SAR目標進行檢測,是目前SAR圖像處理領域研究的熱點問題。視覺注意機制是指人眼對外界環(huán)境中的海量信息進行篩選處理,并只對有用信息做出響應的信息處理機制。視覺注意計算模型是利用數(shù)學建模的方法,結(jié)合計算機提供的仿真環(huán)境,模擬人眼視覺感知系統(tǒng)對信息的處理過程,在圖像分析處理領域,視覺注意機制掀起了研究熱潮。本文將視覺注意機制引入到SAR目標檢測領域,論文的主要研究內(nèi)容和創(chuàng)新之處如下:(1)研究了海內(nèi)外較為成熟的視覺注意計算模型,并與人類視覺感知系統(tǒng)的工作過程相結(jié)合,對應用較廣的Itti經(jīng)典模型和Hou譜殘差模型進行了詳細的分析。闡述了兩種處理方式在SAR圖像處理方面的不足。(2)研究了一種適用于SAR圖像的視覺注意模型。該模型建立在Itti經(jīng)典模型框架的基礎之上,提取能夠表征SAR目標的特征,通過特征整合機制得到SAR圖像的顯著圖。仿真對比實驗的結(jié)果表明,由該模型得到的顯著圖中目標輪廓形狀清晰,定位準確,且計算速度較快。(3)研究了一種基于視覺顯著性的SAR目標快速檢測新算法。該算法首先利用視覺注意模型得到輸入圖像的顯著圖,然后結(jié)合SAR目標的尺寸信息對顯著區(qū)域進行高斯平滑和篩選,然后經(jīng)過感興趣區(qū)域(Region of Interest,ROI)矩形補齊和配準處理,完成對整幅圖像的目標檢測。該方法解決問題的思路較為新穎,結(jié)合了視覺顯著性,仿真實驗結(jié)果表明與傳統(tǒng)的檢測算法相比,該算法具有快速準確的優(yōu)點。
[Abstract]:The work of synthetic Aperture radar (SAR) is not affected by time and weather, and it has strong transmissibility. With these advantages, it plays an important role in the military field. Military operations provide reconnaissance and target detection information needed for military operations. At the same time, they are also widely used in civilian aspects such as weather forecasting, terrain and geomorphology research, road detection, disaster monitoring and other civilian aspects. SAR target detection is based on the grayscale and texture of the target. Shape, edge and direction information in the SAR image to determine the location of the image processing technology to separate the target from the background. With the improvement of the resolution of SAR images and the explosive growth of target information, how to detect SAR targets quickly and effectively is a hot issue in the field of SAR image processing. Visual attention mechanism is an information processing mechanism in which the human eye filters and processes massive information in the external environment and only responds to useful information. Visual attention computing model is to simulate the processing process of human visual perception system by using mathematical modeling method and computer simulation environment. In the field of image analysis and processing, visual attention mechanism has aroused a hot research. In this paper, the visual attention mechanism is introduced into the field of SAR target detection. The main contents and innovations of this paper are as follows: 1) the more mature visual attention computing model at home and abroad is studied and combined with the working process of human visual perception system. The classical Itti model and the Hou spectral residual model are analyzed in detail. The deficiency of two processing methods in SAR image processing is discussed. A visual attention model suitable for SAR image is studied. The model is based on the framework of the classical Itti model and extracts the features that can represent the SAR target. The salient map of the SAR image is obtained by the feature integration mechanism. The simulation results show that the contour of the target is clear, the location is accurate, and the computing speed is faster. (3) A new algorithm for fast detection of SAR targets based on visual saliency is studied. Firstly, the saliency map of the input image is obtained by using visual attention model, then the salient region is smoothed and filtered by Gao Si with the size information of the SAR target, and then the region of interest is patched and registered by rectangle. Complete the target detection of the whole image. The method is novel in solving the problem and combines visual salience. The simulation results show that the algorithm has the advantages of fast and accurate compared with the traditional detection algorithm.
【學位授予單位】：電子科技大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TN957.52

【參考文獻】

相關期刊論文 前1條

1 桑農(nóng),李正龍,張?zhí)煨?人類視覺注意機制在目標檢測中的應用[J];紅外與激光工程;2004年01期

相關博士學位論文 前1條

1 單列;視覺注意機制的若干關鍵技術及應用研究[D];中國科學技術大學;2008年

相關碩士學位論文 前1條

1 陳禮;視覺注意模型的研究及其在ROI圖像壓縮上的應用[D];重慶大學;2008年

，

本文編號：1841544