外文翻譯---一種新的模糊邊緣檢測算法

1、<p>　　A new fuzzy edge detection algorithm</p><p>　　Sun Wei Xia Lianzheng</p><p>　　( Department of Automatic Control Engineering,SoutheastUniversity,Nanjing,210096,China)</p><p&

2、gt;　　Abstract：Based upon the maximum entropy theorem of information theory, a novel fuzzy approach for edge detection is presented .Firstly, a definition of fuzzy partition entropy is proposed after introducing the conce

3、pt of fuzzy probability and fuzzy partition, The relation of the probability partition and the fuzzy c-partition of the image gradient are used in the algorithm。Secondly, based on the conditional probabilities and the fu

4、ry partition, the optimal thresholding is searched adaptively thr</p><p>　　Key words :edge detection ;fuzzy entropy ;image segmentation ;fuzzy partition</p><p>　　Image segmentation is an importa

5、nt topic for image analysis, computer vision and pattern recognition .Until now, many classical edge detection algorithms have been put forward .In recent years, fuzzy set theory has been successfully applied to many are

6、as, such as automation control, image processing, pattern recognition and computer vision, etc .It is generally believed that image processing bears some fuzziness in nature due to the following factors: ①Information los

7、s while mapping 3-D objects </p><p>　　Jin Lizuo .et a1. proposed a new definition of fuzzy partition entropy using the conditional probability and conditional entropy, and designed a new thresholding selecti

8、on algorithm based on the maximum fuzzy entropy .This paper extends the application of the work to the problem of the edge detection and presents a new fuzzy edge detection algorithm .In the algorithm, a gradient image i

9、s considered as being composed of an edge region and a smooth region .Based on the conditional probability and </p><p>　　The rest of this paper is organized as follows .In section 1,we briefly outline the co

10、ncept of fuzzy probability and fuzzy partition entropy .In section 2,we describe the fuzzy edge detection algorithm, In section 3,the experimental results and conclusions are presented.</p><p>　　2.4 Edge det

11、ection</p><p>　　Let the edge image be ,then calculate it as</p><p><b>　　(1)</b></p><p>　　Spurious or weak edges(intensity discontinuities) may result in the image edge r

12、epresentation due to many factors among them are noise and breaks in the boundary between two regions due to non—uniform illumination .In this section, e introduce a simple yet effective procedure for removing spurious o

13、r weak edges .The procedure is as follows:</p><p>　　1)Run a 3×3 pixel window on the edge image .where the center of the window imposed on each location (x, y);</p><p>　　2)Sum the number of

14、points which have been classified as edge in the window, if the number is greater than four, leave these edge points, else they represent weak or spurious edges.</p><p>　　3) Experimental Results and conclusi

15、ons</p><p>　　In this section, the experiments on various kinds of images have been carried out with proposed method .The three original images are selected and shown in Figs.2—4.Fig.2 is an airplane image .T

16、he size of which is 212×200 pixe1.The membership function parameters set(a,b)= (5,157)and the image thresholding is 81.Fig.3 is a baboon image, the size of which is 202×200 pixe1.The membership functions parame

17、ter set(a, b) = (6,164)and the image thresholding is 85.Fig.4 is a Lena image, the size of which</p><p>　　In this paper, we combine conditional probability with fuzzy maximum entropy to introduce a new fuzzy

18、 edge detection algorithm. The experimental resu1ts show that this algorithm performs well. It is verified that segmentation methods, which combine fuzzy statistics. are suitable for theory further research.</p>&

19、lt;p><b>　　作者：孫偉 夏良正</b></p><p><b>　　國籍：中國</b></p><p>　　出處：東南大學(xué)學(xué)報(英文版).第二期.卷20.2003.</p><p>　　一種新的模糊邊緣檢測算法</p><p><b>　　孫偉 夏良正</b>&

20、lt;/p><p>　　(東南大學(xué)自動控制系，南京210096)</p><p>　　摘要:基于信息論中最大熵原理，提出了一種新的模糊邊緣檢測算法。首先介紹了模糊概率、用條件概率與條件熵定義模糊劃分熵的概念以及模糊劃分的原理。算法利用了自然劃分以及梯度圖像模糊劃分的關(guān)系，在條件概率與模糊劃分熵的基礎(chǔ)上，通過最大模糊熵原則實現(xiàn)圖像分割中最優(yōu)閾值的自動提取，從而實現(xiàn)圖像的邊緣檢測。對不同測試圖像的

21、邊緣檢測結(jié)果進行比較，表明了該算法的有效性。</p><p>　　關(guān)鍵詞：邊緣檢測；模糊熵；圖像分割；模糊劃分</p><p>　　圖像分割是圖像分析、計算機視覺和模式識別的一個重要課題，直至現(xiàn)在,許多經(jīng)典的邊緣檢測算法已提出。在最近的幾年，模糊理論已經(jīng)在很多領(lǐng)域運用，如自動化控制,圖象處理,模式識別和計算機視覺等。人們普遍認為在自然條件下圖像處理負有一定的模糊性是由于下列因素：①信息的丟

22、失當3-D的物體轉(zhuǎn)換為2-D圖像；②在定義上(例如邊緣,邊界地區(qū),紋理,等等)的二異性；③低級別圖像處理后解釋的二異性因此，模糊技術(shù)經(jīng)常在圖像分割上使用。</p><p>　　利用條件概率與條件熵，Jin Lizuo等人提出了一個新的定義模糊劃分熵，并設(shè)計了一個新的閾值選擇算法基于最大模糊熵。本文把該方法延伸運用到邊緣檢測方法，并提出了一個新的模糊邊緣檢測算法。在算法中,梯度圖像被視為是由一個邊緣地區(qū)的和平穩(wěn)的地

23、區(qū)組成的，基于條件概率及模糊劃分熵。最優(yōu)閾值搜尋是模糊最大熵原則實現(xiàn)的。這有兩個主要的區(qū)別問題的邊緣檢測和圖像分割。首先，這一問題實際上是減少了兩個級別的閾值問題,目的是通過閾值把圖像分割成兩個區(qū)域:一個邊緣地區(qū)和平穩(wěn)地區(qū)。第二,通過梯度圖像處理，可以找到最好的緊湊的有代表性的圖像邊緣和輪廓。實驗結(jié)果表明該算法的有效性。文章的其余部分組織如下：第一部分，我們簡述概念模糊概率及模糊劃分。第二部分,我們描述的模糊邊緣檢測算法。在第三部分中，

24、實驗結(jié)果和結(jié)論等。</p><p><b>　　2.4 邊緣檢測</b></p><p>　　另邊緣圖像為e(x, y)，然后計算如下：</p><p><b>　　(1) </b></p><p>　　由于其他很多的原因，邊沿會變成假的或效果不好的（強度不連續(xù)）；其中有噪音和兩部分邊界斷裂是,

25、因非均勻照明。在這部分中,我們引進了一種簡單而有效的程序,清除假的或效果不好的邊沿。程序如下: </p><p>　　1)在邊緣圖像上，運行一個3×3像素的窗口,窗口的中心在點(x, y)上; </p><p>　　2) 把在窗口中，已經(jīng)通過邊緣分類的點全部加起來，假如這個數(shù)字大于4，離開這個邊緣點，否則他們代表假的或效果不好的邊緣點。</p><p>

26、　　3) 實驗結(jié)果與結(jié)論</p><p>　　在這一部分中，所有的實驗都是在前面提出的方法上處理。三幅原始圖像和處理過的圖像如圖2~4所示。2是一幅飛機的圖像，大小為212×200個像素點，成員函數(shù)的參數(shù)設(shè)定(a, b)=(5,157)和圖像閾值是81。3是狒狒的圖像，大小為202×200個像素點 ，成員函數(shù)的參數(shù)設(shè)定(a, b)=(6,164)和圖像閾值是85。4是Lena的圖像，大小為21