基于神經(jīng)網(wǎng)絡(luò)的玻璃缺陷類型識(shí)別方法畢業(yè)設(shè)計(jì)(已修改)
2025-07-30 11:40 本頁(yè)面
【正文】 中北大學(xué) 20xx 屆畢業(yè)論文 畢業(yè)論文 基于神經(jīng)網(wǎng)絡(luò)的玻璃缺陷類型識(shí)別方法 學(xué)生姓名： 毛睿達(dá) 學(xué)號(hào)： 1105064125 學(xué) 院： 信息與通信工程學(xué)院 專 業(yè)： 電子信息工程 指導(dǎo)教師： 金永 20xx 年 6 月 中北大學(xué) 20xx 屆畢業(yè)論文 基于神經(jīng)網(wǎng)絡(luò)的玻璃缺陷類型識(shí)別方法 摘要 在玻璃生產(chǎn)過(guò)程中，由于受工藝和環(huán)境限制，會(huì)產(chǎn)生各種缺陷，這些缺陷不僅影響了玻璃制品的外觀質(zhì)量，也降低了玻璃的 使用價(jià)值和再次加工率。為了提高玻璃質(zhì)量及方便玻璃質(zhì)量等級(jí)劃分，必須對(duì)缺陷進(jìn)行分類。本文針對(duì)玻璃缺陷圖像的特點(diǎn)，基于圖像處理與模式識(shí)別技術(shù)，研究了缺陷自動(dòng)分類算法，替代了傳統(tǒng)的人工分類方法，提高了分類的精度和效率。 本文首先分析了缺陷圖像的噪聲類型及特點(diǎn)，采用中值濾波算法對(duì)缺陷圖像進(jìn)行降噪處理，消除了各種噪聲干擾 。然后針對(duì)缺陷圖像邊緣的灰度變化特點(diǎn)，基于邊緣檢測(cè)技術(shù)，較為完整的提取出了目標(biāo)缺陷的核心輪廓，完成了圖像的預(yù)處理。 在預(yù)處理的基礎(chǔ)上，根據(jù)各類缺陷在形狀上的差異，利用 Hu 不變矩提取出了缺陷的形狀特征， 并驗(yàn)證了其抗平移、抗旋轉(zhuǎn)性，將 Hu不變矩提取出來(lái)的 7 個(gè)特征值作為缺陷分類器的輸入向量。 為了區(qū)分不同類型的缺陷，研究了 感知器 神經(jīng)網(wǎng)絡(luò)分類器的設(shè)計(jì)，設(shè)計(jì)了感知器神經(jīng)網(wǎng)絡(luò)的算法。最后，通過(guò)實(shí)驗(yàn)整體驗(yàn)證了缺陷分類算法的有效性，取得了良好的識(shí)別效果，為后期地投入實(shí)際生產(chǎn)打下了堅(jiān)實(shí)的基礎(chǔ)。 關(guān)鍵詞 : 玻璃缺陷，圖像預(yù)處理，特征提取，神經(jīng)網(wǎng)絡(luò) 中北大學(xué) 20xx 屆畢業(yè)論文 Identification method of glass defect type based on Neural Network Abstract In the glass production process, due to the technological and environmental restrictions, will produce a variety of defects, these defects not only affects the appearance quality of glass products, but also reduces the value of the use of glass and re processing rate. In order to improve the quality of glass and glass quality grades, we must classify the defects.. The according to the characteristics of glass defect image, based on image processing and pattern recognition technology of automatic defect classification algorithm, replacing the traditional manual classification method, improves the classification accuracy and efficiency. This paper first analyzes the noise type and character of the defect image, the median filtering algorithm for reduction of defect image, eliminating the noise。 then in accordance with the characteristics of gray level of image edge defects, based on the edge detection technique, more plete extraction of the core dimensions of the defect, pleted the image preprocessing. On the basis of preprocessing, according to the difference between the various types of defects in shape using Hu invariant moments to extract the features of the shape of the defects, and verified the robustness against translation, anti rotation, the Hu invariant moments to extract out of seven feature values as the input vectors of the defect classifier. In order to distinguish between different types of defects of perceptron neural work classifier design, design the perceptron neural work algorithm. Finally, the whole experiment to verify the effectiveness of defect classification algorithm, and achieved good recognition effect, late into the actual production and lay a solid foundation. Keywords: glass flaw, image preprocessing, feature extraction, neural work 中北大學(xué) 20xx 屆畢業(yè)論文 1 引言 ......................................................................................................................................... 5 課題背景及意義 .............................................................................................................. 5 國(guó)內(nèi)外研究現(xiàn)狀 .............................................................................................................. 5 課題研究?jī)?nèi)容以及論文安排 ............................................................................................. 7 2 玻璃缺陷圖像的預(yù)處理 ............................................................................................................. 8 2. 1 玻璃缺陷圖像噪聲分析 ................................................................................................... 8 攝像機(jī)產(chǎn)生的噪聲 ..................................................................................................... 9 圖像信號(hào)數(shù)字化產(chǎn)生的像素抖動(dòng) ...............................................................................10 2 .2 玻璃缺陷圖像噪聲消除 .................................................................................................. 11 玻璃缺陷圖像的分割 ......................................................................................................13 3 玻璃缺陷圖 像特征提取 ............................................................................................................18 3. 1 缺陷特征選取的原則 ......................................................................................................18 3. 2 基于 Hu 不變矩特征的提取 ............................................................................................18 Hu 不變矩概述 .......................................................................................................18 基于 Hu 不變矩的特征提取實(shí)驗(yàn)及結(jié)果 .....................................................................22 4 基于感知器神經(jīng)網(wǎng)絡(luò)的玻璃缺陷分類識(shí)別 ...........................................................................................22 感知器神經(jīng)網(wǎng)絡(luò)簡(jiǎn)介 ......................................................................................................24 玻璃缺陷類型識(shí)別及結(jié)果分析 ........................................................................................25 輸入、輸出層神經(jīng)元個(gè)數(shù) .........................................................................................25 玻璃缺陷類型識(shí)別及結(jié)果分析 ..................................................................................25 5 結(jié)論 ........................................................................................
研究報(bào)告相關(guān)推薦
鄂ICP備17016276號(hào)-1