【正文】
【精品】畢業(yè)設(shè)計(jì)(論文) 灰度圖像特征提取算法研究 摘 要 圖像特征提取一直是圖像研究和計(jì)算機(jī)識(shí)別中一個(gè)值得探討的問(wèn)題在計(jì)算機(jī)科學(xué)醫(yī)療輔助診斷軍事工業(yè)測(cè)量等各種領(lǐng)域都正在發(fā)揮著越來(lái)越重要的作用尤其是醫(yī)療檢索系統(tǒng)中如何準(zhǔn)確定位和提取關(guān)鍵特征往往是首要解決的問(wèn)題之一是提高醫(yī)療儀器識(shí)別率的關(guān)鍵在醫(yī)療檢索系統(tǒng)中準(zhǔn)確的定位和提取圖像的邊緣特征能夠使醫(yī)生快速準(zhǔn)確的判別出病因本文針對(duì)邊緣特征提取算法和灰度直方圖特征提取算法進(jìn)行了特征提取試驗(yàn)與分析 邊緣特征提取本文通 過(guò)采取 Robets 算法 Sobel 算法 Canny 算法小波變換算法對(duì)一幅圖像做了不同的邊緣特征提取結(jié)果表明本文采取的 Canny 邊緣特征提取算子法明顯優(yōu)于其他邊緣特征提取算法且能夠最為接近的描述灰度圖像的邊緣特征信息 灰度直方圖特征對(duì)于邊緣算子法提取出的灰度邊緣特征圖像通過(guò)人眼觀(guān)測(cè)有時(shí)可能會(huì)觀(guān)測(cè)不到明顯的區(qū)別本文通過(guò)分析其灰度灰度直方圖特征提取其中的灰度特征量灰度均值 Mean方差 Variance熵 ntropy能量 Energy根據(jù)特征量的變化情況分析圖像邊緣提取效果的差異 本算法編程采用 C語(yǔ)言在 Visual studio2021平臺(tái)下實(shí)現(xiàn)通過(guò)各種算法對(duì)圖像特征提取的描述實(shí)現(xiàn)其不同的提取效果結(jié)果顯示 C 語(yǔ)言能更好的表示各種不同算法在圖像處理中的編程 關(guān)鍵詞灰度圖像特征提取邊緣特征灰度直方圖特征量 Abstract Research and puter image feature extraction has been a problem that worth exploring in visual identification In puter science medical diagnosis military industry measurement and so on various fields are playing an increasingly important role especially in medical retrieval system how t accurately locate and extract the key features are often the first to solve one of the problems Is the key to improve medical instrument recognition rate In medical retrieval system accurately locate and extract the image edge features Can make the doctor quickly accurately identify the cause Based on edge feature extraction algorithm and the algorithm for calculating grayscale histogram feature extraction feature extraction experiment and analyzed The edge feature extraction Algorithm in this paper by adopting Robets Sobel algorithm the Canny algorithm the wavelet transform algorithm to an image made of different edge feature extraction Results show that the adopted Canny edge operator of feature extraction method is superior to other edge feature extraction algorithms And can describe the most close to the edge of gray image feature information Gray level histogram feature for gray level edge operator method to extract the edge character images by the human eye observation may sometimes observed no significant difference between the two In this article through analyzing the characteristic of the gray level histogram grayscale andExtract the grayscale characteristics grayscale average mean and variance variance entropy entropy energy energy according to the change of characteristic analysis of image edge extraction effect difference The algorithm programming using c language in Visual studio2021 platform through a variety of algorithms for image features extraction and description of the effect of different extraction results show that the c language is better said all sorts of different algorithms in image processing programming Keywords Gray image Feature extraction Edge features Gray levelhistogram characteristic 目 錄 第 1 章 緒論 1 11 研究背景及意義 1 12 論文的研究?jī)?nèi)容 1 13 本文的結(jié)構(gòu)安排 2 第 2 章 邊緣圖像預(yù)處理技術(shù) 3 21 銳化 3 22 彩色圖像灰度化 4 23 圖像二值化 5 24 本章小結(jié) 5 第 3 章 圖像邊緣特征提取算法 6 31 邊緣特征提取步驟 6 32 傳統(tǒng)邊緣特征提取主要算法及其效果分析 7 Roberts 邊緣算子 7 Sobel 邊緣算子 8 Prewitt 算子 10 Laplace 算子 11 33 改進(jìn)的邊緣特征提取算法及其效果分析 13 Canny 邊緣檢測(cè)子算法 13 小波變換邊緣檢測(cè)算法 17 34 算法比較 21 35 本章小結(jié) 23 第 4 章 灰度直方圖特征提取算法 24 41 灰度直方圖的定義 24 42 基于直方圖的特征統(tǒng)計(jì) 24 43 灰度直方圖特征提取的 C 實(shí)現(xiàn)代碼及直方圖顯示 25 44 統(tǒng)計(jì)量描述直方圖 27 45 本章小結(jié) 29 第 5 章 總結(jié)與展望 30 51 總結(jié) 30 52 展望 30 參考文獻(xiàn) 31 致 謝 32 第 1 章 緒論 11 研究背景及意義 由于醫(yī)療特征成像檢測(cè)系統(tǒng)診斷方式與其他診斷方式相比可以快速準(zhǔn)確的查找病 因并且不會(huì)對(duì)病人造成太大的傷害因此成像診斷技術(shù)已成為目前發(fā)展最為迅速的診斷技術(shù)通過(guò)醫(yī)療檢測(cè)病人器官成像特征是一種有效的檢測(cè)方式但是現(xiàn)實(shí)中對(duì)圖像特征提取的技術(shù)還不夠完善雖然人們?cè)诖嘶A(chǔ)上研究出了一系列的特征及其提取算法但是隨著計(jì)算機(jī)技術(shù)和成像水平的高速發(fā)展現(xiàn)有的技術(shù)已經(jīng)無(wú)法滿(mǎn)足醫(yī)療診斷的需要本文主要對(duì)當(dāng)前的邊緣級(jí)灰度直方圖的特征提取算法進(jìn)行了研究分析 圖像邊緣特征提取涉及圖像中對(duì)象圖片的特征提取即怎樣識(shí)別圖像中物體的輪廓是數(shù)字圖像分析處理的前提邊緣特征提取結(jié)果直接影響著圖