畢業(yè)論文 基于小波變換的圖像邊緣檢測技術(shù)研究.doc
約22頁DOC格式手機打開展開
畢業(yè)論文 基于小波變換的圖像邊緣檢測技術(shù)研究,摘要：圖像的邊緣在圖像識別，分割，增強以及壓縮等領(lǐng)域中都有較廣泛的應(yīng)用，也是它們的基礎(chǔ)。邊緣檢測是圖像處理與分析中最基礎(chǔ)的內(nèi)容之一，也是至今仍沒有圓滿解決的一類問題。圖像的邊緣包含了圖像的位置、輪廓燈特征，是圖像的基本特征之一，廣泛的應(yīng)用于特征描述、圖像復(fù)原、增強、壓縮和處理中。因此，圖像邊緣和輪廓特征的檢測與提取方法...
內(nèi)容介紹
此文檔由會員 ljjwl8321 發(fā)布
摘 要：圖像的邊緣在圖像識別，分割，增強以及壓縮等領(lǐng)域中都有較廣泛的應(yīng)用，也是它們的基礎(chǔ)。邊緣檢測是圖像處理與分析中最基礎(chǔ)的內(nèi)容之一，也是至今仍沒有圓滿解決的一類問題。
圖像的邊緣包含了圖像的位置、輪廓燈特征，是圖像的基本特征之一，廣泛的應(yīng)用于特征描述、圖像復(fù)原、增強、壓縮和處理中。因此，圖像邊緣和輪廓特征的檢測與提取方法，一直是圖像處理與分析技術(shù)中的研究熱點。
本文研究了一些邊緣檢測算法，包括Canny算法、Sobel算法、Roberts算法、Prewitt算法和拉普拉斯算法等。傳統(tǒng)的圖像邊緣檢測算法對受到噪聲污染的圖像效果很差，解決該問題的主要方法就是設(shè)置閾值，把得到的圖像高頻部分與閾值相比較以達到去噪的目的。
論文的主要目的是進行圖像邊緣檢測算法性能比較的研究。實驗結(jié)果表明，相比傳統(tǒng)的檢測算法，小波變換具有很大的優(yōu)勢。
關(guān)鍵詞：邊緣檢測， 圖像處理， 檢測算法
Title：Image Edge Detection Based on Wavelet Transformation
Abstract: The image border is in pattern recognition , division , has broader application , is also their basis in fields such as strengthening and compressing. The border detecting is image treatment and one of the basis content, is also a kind of problem not being brought to a satisfactory settlement so far still most in analysis.
The image border has contained the image location , the outline light characteristic , has been one of the image essential features , broad applying to the characteristic describes that , the image restores , the treatment strengthening , compressing a sum is hit by. Therefore, image border composes in reply outline characteristic detecting with extracting method, the hot spot studying in being always that the image handles and analyses a technology.
The algorithm having studied some border detect algorithm , having included the Canny algorithm , Sobel algorithm , Roberts algorithm , Prewitt algorithm and Laplace waits for the main body of a book. The tradition image border detect algorithm is dispatched face to face very much by the image effect that noise contaminates, the method resolving main part that a problem is to interpose threshold value , compare the image high frequency part and threshold value to achieve go and chirp purpose.
The thesis major objective is to carry out parallel research of image border detect algorithm function. The experiment bear fruit is indicated , is compared with each other the tradition detect algorithm , minor wave alternation have very big advantage.
Keywords: Border detecting, Image treatment , Detect algorithm
目 錄
摘 要 I
Abstract II
緒 論 1
1 圖像邊緣檢測概述 2
1.1圖像邊緣檢測的發(fā)展前景 2
1.2圖像邊緣檢測的應(yīng)用 2
2 基于一階微分的邊緣檢測算法 4
2．1 Roberts算子 4
2.2 Soble 算子 4
2.3 Prewitt算子 4
3 基于二階微分的邊緣檢測算法 6
3.1 Laplacian 算子 6
3.2 Log 算子 6
3.3 Canny 算子 6
4 小波變換的邊緣檢測算法 8
4.1小波變換與多尺度邊緣檢測 9
4.2 數(shù)字圖像的小波變換 10
4.3 小波變換 10
5 基于Matlab的實驗結(jié)果與分析 12
5.1 Matlab簡介 12
5.2 小波變換的實驗結(jié)果與分析 12
結(jié) 論 13
注 釋 14
參考文獻 15
致 謝 16
圖像的邊緣包含了圖像的位置、輪廓燈特征，是圖像的基本特征之一，廣泛的應(yīng)用于特征描述、圖像復(fù)原、增強、壓縮和處理中。因此，圖像邊緣和輪廓特征的檢測與提取方法，一直是圖像處理與分析技術(shù)中的研究熱點。
本文研究了一些邊緣檢測算法，包括Canny算法、Sobel算法、Roberts算法、Prewitt算法和拉普拉斯算法等。傳統(tǒng)的圖像邊緣檢測算法對受到噪聲污染的圖像效果很差，解決該問題的主要方法就是設(shè)置閾值，把得到的圖像高頻部分與閾值相比較以達到去噪的目的。
論文的主要目的是進行圖像邊緣檢測算法性能比較的研究。實驗結(jié)果表明，相比傳統(tǒng)的檢測算法，小波變換具有很大的優(yōu)勢。
關(guān)鍵詞：邊緣檢測， 圖像處理， 檢測算法
Title：Image Edge Detection Based on Wavelet Transformation
Abstract: The image border is in pattern recognition , division , has broader application , is also their basis in fields such as strengthening and compressing. The border detecting is image treatment and one of the basis content, is also a kind of problem not being brought to a satisfactory settlement so far still most in analysis.
The image border has contained the image location , the outline light characteristic , has been one of the image essential features , broad applying to the characteristic describes that , the image restores , the treatment strengthening , compressing a sum is hit by. Therefore, image border composes in reply outline characteristic detecting with extracting method, the hot spot studying in being always that the image handles and analyses a technology.
The algorithm having studied some border detect algorithm , having included the Canny algorithm , Sobel algorithm , Roberts algorithm , Prewitt algorithm and Laplace waits for the main body of a book. The tradition image border detect algorithm is dispatched face to face very much by the image effect that noise contaminates, the method resolving main part that a problem is to interpose threshold value , compare the image high frequency part and threshold value to achieve go and chirp purpose.
The thesis major objective is to carry out parallel research of image border detect algorithm function. The experiment bear fruit is indicated , is compared with each other the tradition detect algorithm , minor wave alternation have very big advantage.
Keywords: Border detecting, Image treatment , Detect algorithm
目 錄
摘 要 I
Abstract II
緒 論 1
1 圖像邊緣檢測概述 2
1.1圖像邊緣檢測的發(fā)展前景 2
1.2圖像邊緣檢測的應(yīng)用 2
2 基于一階微分的邊緣檢測算法 4
2．1 Roberts算子 4
2.2 Soble 算子 4
2.3 Prewitt算子 4
3 基于二階微分的邊緣檢測算法 6
3.1 Laplacian 算子 6
3.2 Log 算子 6
3.3 Canny 算子 6
4 小波變換的邊緣檢測算法 8
4.1小波變換與多尺度邊緣檢測 9
4.2 數(shù)字圖像的小波變換 10
4.3 小波變換 10
5 基于Matlab的實驗結(jié)果與分析 12
5.1 Matlab簡介 12
5.2 小波變換的實驗結(jié)果與分析 12
結(jié) 論 13
注 釋 14
參考文獻 15
致 謝 16
