機(jī)器視覺與圖像處理5篇報(bào)告

1、機(jī)器視覺與圖像處理實(shí)驗(yàn)報(bào)告姓名：張鵬濤 班級(jí)：自動(dòng)1302 學(xué)院：電氣工程學(xué)院 學(xué)號(hào)：201323020219 任課老師：石慶升 時(shí)間：2016/04/14 實(shí)驗(yàn)項(xiàng)目1：圖像處理軟件的安裝與應(yīng)用一、實(shí)驗(yàn)?zāi)康?#160; 1、熟悉并掌握MATLAB工具的安裝及界面結(jié)構(gòu)；2、實(shí)現(xiàn)圖像的讀取、顯示、對(duì)比度提高，實(shí)現(xiàn)圖片磁盤文件的寫入等操作。二、實(shí)驗(yàn)環(huán)境MATLAB R2008以上版本、WIN7計(jì)算機(jī)三、實(shí)驗(yàn)內(nèi)容 1.利用提供的安裝源文件，實(shí)現(xiàn)Matlab軟件的安裝 2.安裝完成后，觀察啟動(dòng)界面的結(jié)構(gòu)組成 3.讀取和顯示圖片>>I = imread('rice.png');

2、>>imshow(I) 4. 使用形態(tài)學(xué)開估計(jì)背景>>background = imopen(I,strel('disk',15);>> figure, surf(double(background(1:8:end,1:8:end),zlim(0 255);>> set(gca,'ydir','reverse'); 5. 從原始圖像減去Backround形象>> I2 = imsubtract(I,background);>> imshow(I2) 6. 增加圖像的對(duì)比>

3、;> I3 = imadjust(I2);>> imshow(I3); 7. 閾值圖像>> level = graythresh(I3);>>bw = im2bw(I3,level);>>imshow(bw) 8. 標(biāo)簽圖像中的對(duì)象 9. 檢查標(biāo)簽的矩陣>> rect = 105 125 10 10;>>grain = imcrop(labeled,rect) 10. 整個(gè)標(biāo)簽視圖矩陣>> RGB_label = label2rgb(labeled, spring, 'c', 's

4、huffle');>>imshow(RGB_label) 11. 測(cè)量圖像中對(duì)象屬性>> graindata = regionprops(labeled,'basic')>> graindata(50).Area 12. 計(jì)算圖像中對(duì)象的統(tǒng)計(jì)特性>> allgrains = graindata.Area;>>max_area = max(allgrains)>> biggrain = find(allgrains=max_area)>> mean(allgrains) 13. 創(chuàng)建一個(gè)顆

5、粒的面積的柱狀圖>> nbins = 20;>>figure,hist(allgrains,nbins)四、實(shí)驗(yàn)總結(jié)1）安裝過程中需要序列號(hào)，這個(gè)在百度上搜plp找到相應(yīng)版本的的序列號(hào)就可以了，安裝的過程中可能有多個(gè)文件包吧，先安裝第一個(gè)，在第一個(gè)安裝好以后，程序會(huì)提醒你下一個(gè)的安裝目錄，你找到相應(yīng)文件夾引導(dǎo)一下即可，其他的相同。 最后還要告訴你一下，win7 與matlab 7.0可能不兼容，在安裝好的圖標(biāo)上右擊屬性，選擇高級(jí)，將以兼容性運(yùn)行該程序的對(duì)勾消掉即可。2）MATLAB工具箱有著豐富的圖像工具，包括灰度處理，對(duì)比度增強(qiáng)，形狀識(shí)別處理等。實(shí)驗(yàn)項(xiàng)目2：圖像增強(qiáng)技

6、術(shù)應(yīng)用一、實(shí)驗(yàn)?zāi)康?#160; 1、熟悉并掌握MATLAB圖像處理工具箱的使用；2、理解并掌握常用的圖像的增強(qiáng)技術(shù)。二、實(shí)驗(yàn)環(huán)境MATLAB R2008以上版本、WIN7計(jì)算機(jī)三、實(shí)驗(yàn)內(nèi)容1. 加載圖片>> pout = imread('pout.tif');>> tire = imread('tire.tif');>> X map = imread('shadow.tif');>> shadow = ind2rgb(X,map); 2. 調(diào)整圖像>> width = 210;>&

7、gt;images = pout, tire, shadow;for k = 1:3 dim = size(imagesk); imagesk = imresize(imagesk,width*dim(1)/dim(2) width,'bicubic');end>>pout = images1;>>tire = images2;>>shadow = images3; 3. 提高灰度圖像>>pout_imadjust = imadjust(pout);>>pout_histeq = histeq(pout);>&g

8、t;pout_adapthisteq = adapthisteq(pout);>>imshow(pout);>>title('Original');>>figure, imshow(pout_imadjust);>>title('Imadjust');>>figure, imshow(pout_histeq);>>title('Histeq');>>figure, imshow(pout_adapthisteq);>>title('Adapth

9、isteq');>>tire_imadjust = imadjust(tire);>>tire_histeq = histeq(tire);>>tire_adapthisteq = adapthisteq(tire);>>figure, imshow(tire);>>title('Original');>>figure, imshow(tire_imadjust);>>title('Imadjust');>>figure, imshow(tire_histeq

10、);>>title('Histeq');>>figure, imshow(tire_adapthisteq);>>title('Adapthisteq'); >> figure, imhist(pout), title('pout.tif');>>figure, imhist(tire), title('tire.tif'); 4. 提高彩色圖像>> srgb2lab = makecform('srgb2lab');>>lab2sr

11、gb = makecform('lab2srgb');>>shadow_lab = applycform(shadow, srgb2lab); >> max_luminosity = 100;>>L = shadow_lab(:,:,1)/max_luminosity;>> shadow_imadjust = shadow_lab;>> shadow_imadjust(:,:,1) = imadjust(L)*max_luminosity;>>shadow_imadjust = applycform(sha

12、dow_imadjust, lab2srgb);>>shadow_histeq = shadow_lab;>>shadow_histeq(:,:,1) = histeq(L)*max_luminosity;>>shadow_histeq = applycform(shadow_histeq, lab2srgb);>>shadow_adapthisteq = shadow_lab;>>shadow_adapthisteq(:,:,1) = adapthisteq(L)*max_luminosity;>>shadow_adap

13、thisteq = applycform(shadow_adapthisteq, lab2srgb);>>figure, imshow(shadow);>>title('Original');>>figure, imshow(shadow_imadjust);>>title('Imadjust');>> figure, imshow(shadow_histeq);>>title('Histeq');>>figure, imshow(shadow_adapthist

14、eq);>>title('Adapthisteq');四、實(shí)驗(yàn)總結(jié)直方圖均衡化是圖像處理領(lǐng)域中利用圖像直方圖對(duì)對(duì)比度進(jìn)行調(diào)整的方法。這種方法通常用來增加許多圖像的局部對(duì)比度，尤其是當(dāng)圖像的有用數(shù)據(jù)的對(duì)比度相當(dāng)接近的時(shí)候。通過這種方法，亮度可以更好地在直方圖上分布。這樣就可以用于增強(qiáng)局部的對(duì)比度而不影響整體的對(duì)比度，直方圖均衡化通過有效地?cái)U(kuò)展常用的亮度來實(shí)現(xiàn)這種功能。實(shí)驗(yàn)項(xiàng)目3：圖像變換技術(shù)應(yīng)用一、實(shí)驗(yàn)?zāi)康?#160; 1、熟悉并掌握MATLAB圖像處理工具箱的使用；2、理解并掌握常用的仿射變換、非反射性變換、投影變換等圖像變換技術(shù)。二、實(shí)驗(yàn)環(huán)境MATLAB R200

15、8以上版本、WIN7計(jì)算機(jī)三、實(shí)驗(yàn)內(nèi)容 1)創(chuàng)建棋盤圖I = checkerboard(10,2);imshow(I)title('original')scale = 1.2; angle = 40*pi/180; tx = 0; ty = 0; sc = scale*cos(angle);ss = scale*sin(angle);T = sc -ss; ss sc; tx ty;t_nonsim = maketform('affine',T);I_nonreflective_similarity = imtransform(I,t_nonsim,'F

16、illValues',.3);figure, imshow(I_nonreflective_similarity);title('nonreflective similarity')I_nonreflective_similarity,xdata, ydata = imtransform(I,t_nonsim,'FillValues',.3);figure, imshow(I_nonreflective_similarity,'XData',xdata,'YData',ydata)axis ontitle('non

17、reflective similarity')2)應(yīng)用無反射相似棋盤scale = 1.5; angle = 10*pi/180; tx = 0; ty = 0; a = -1; sc = scale*cos(angle);ss = scale*sin(angle);T = sc -ss; a*ss a*sc; tx ty;t_sim = maketform('affine',T);I_similarity,xdata,ydata = imtransform(I,t_sim,'FillValues',.3);figure, imshow(I_simila

18、rity,'XData',xdata,'YData',ydata)axis ontitle('similarity')3)仿射變換應(yīng)用于棋盤T = 1 0.3; 1 1; 0 0;t_aff = maketform('affine',T);I_affine = imtransform(I,t_aff,'FillValues',.3);figure, imshow(I_affine)title('affine')4)射影變換應(yīng)用于棋盤T = 1 0 0.008; 1 1 0.01; 0 0 1;t_p

19、roj = maketform('projective',T);I_projective = imtransform(I,t_proj,'FillValues',.3);figure, imshow(I_projective)title('projective')5)多項(xiàng)式變換應(yīng)用于棋盤xybase = reshape(randn(12,1),6,2);t_poly = cp2tform(xybase,xybase,'polynomial',2);T = 0 0; 1 0; 0 1; 0.001 0; 0.02 0; 0.01 0

20、;t_poly.tdata = T;I_polynomial = imtransform(I,t_poly,'FillValues',.3);figure, imshow(I_polynomial)title('polynomial')6)分段線性變換應(yīng)用于棋盤xy_input = 10 10; 10 30; 30 30; 30 10;xy_base = 10 10; 10 30; 40 35; 30 10;t_piecewise_linear = cp2tform(xy_input,xy_base,'piecewise linear');I_p

21、iecewise_linear = imtransform(I,t_piecewise_linear);figure, imshow(I_piecewise_linear)title('piecewise linear')7)正弦變換應(yīng)用于棋盤nrows,ncols = size(I);xi,yi = meshgrid(1:ncols,1:nrows);a1 = 5; a2 = 3;imid = round(size(I,2)/2);u = xi + a1*sin(pi*xi/imid);v = yi - a2*sin(pi*yi/imid);tmap_B = cat(3,u,

22、v);resamp = makeresampler('linear','fill');I_sinusoid = tformarray(I,resamp,2 1,1 2,tmap_B,.3);figure, imshow(I_sinusoid)title('sinusoid')8)桶轉(zhuǎn)換適用于棋盤xt = xi(:) - imid;yt = yi(:) - imid;theta,r = cart2pol(xt,yt);a = .001; s = r + a*r.3;ut,vt = pol2cart(theta,s);u = reshape(ut,s

23、ize(xi) + imid;v = reshape(vt,size(yi) + imid;tmap_B = cat(3,u,v);I_barrel = tformarray(I,resamp,2 1,1 2,tmap_B,.3);figure, imshow(I_barrel)title('barrel')9)插針轉(zhuǎn)換應(yīng)用到棋盤xt = xi(:) - imid;yt = yi(:) - imid;theta,r = cart2pol(xt,yt);a = -.0005; % Try varying the amplitude of the cubic term.s = r

24、+ a*r.3;ut,vt = pol2cart(theta,s);u = reshape(ut,size(xi) + imid;v = reshape(vt,size(yi) + imid;tmap_B = cat(3,u,v);I_pin = tformarray(I,resamp,2 1,1 2,tmap_B,.3);figure, imshow(I_pin)title('pin cushion')figuresubplot(5,2,1),imshow(I),title('original')subplot(5,2,2),imshow(I_nonrefle

25、ctive_similarity),title('nonreflective similarity')subplot(5,2,3),imshow(I_similarity),title('similarity')subplot(5,2,4),imshow(I_affine),title('affine')subplot(5,2,5),imshow(I_projective),title('projective')subplot(5,2,6),imshow(I_polynomial),title('polynomial

26、9;)subplot(5,2,7),imshow(I_piecewise_linear),title('piecewise linear')subplot(5,2,8),imshow(I_sinusoid),title('sinusoid')subplot(5,2,9),imshow(I_barrel),title('barrel')subplot(5,2,10),imshow(I_pin),title('pin cushion')四、實(shí)驗(yàn)總結(jié)為了用正交函數(shù)或正交矩陣表示圖像而對(duì)原圖像所作的二維線性可逆變換。一般稱原始圖像為空間域

27、圖像，稱變換后的圖像為轉(zhuǎn)換域圖像，轉(zhuǎn)換域圖像可反變換為空間域圖像。圖像處理中所用的變換都是酉變換，即變換核滿足正交條件的變換。經(jīng)過酉變換后的圖像往往更有利于特征抽取、增強(qiáng)、壓縮和圖像編碼。 實(shí)驗(yàn)項(xiàng)目4：圖像特征提取技術(shù)應(yīng)用一、實(shí)驗(yàn)?zāi)康?#160; 1、熟悉并掌握MATLAB圖像處理工具箱的使用；2、理解并掌握常用的圖像特征提取技術(shù)。二、實(shí)驗(yàn)環(huán)境MATLAB R2008以上版本、WIN7計(jì)算機(jī)三、實(shí)驗(yàn)內(nèi)容 以圓的測(cè)量為例：1.讀入圖像；RGB = imread('pillsetc.png');imshow(RGB);2.圖形閾值處理；I = rgb2gray(RGB);thres

28、hold = graythresh(I);bw = im2bw(I,threshold);imshow(bw)3.消除圖像噪聲；bw = bwareaopen(bw,30);se = strel('disk',2);bw = imclose(bw,se);bw = imfill(bw,'holes');imshow(bw)4.尋找邊界；B,L = bwboundaries(bw,'noholes');imshow(label2rgb(L, jet, .5 .5 .5)hold onfor k = 1:length(B) boundary = Bk

29、; plot(boundary(:,2), boundary(:,1), 'w', 'LineWidth', 2)end5. 確定哪些對(duì)象是圓的stats = regionprops(L,'Area','Centroid');threshold = 0.94;for k = 1:length(B) boundary = Bk; delta_sq = diff(boundary).2; perimeter = sum(sqrt(sum(delta_sq,2); area = stats(k).Area; metric = 4*pi*

30、area/perimeter2; metric_string = sprintf('%2.2f',metric); if metric > threshold centroid = stats(k).Centroid; plot(centroid(1),centroid(2),'ko'); endtext(boundary(1,2)-35,boundary(1,1)+13,metric_string,'Color','y',. 'FontSize',14,'FontWeight','b

31、old');endtitle('Metrics closer to 1 indicate that ',.'the object is approximately round'); 四、實(shí)驗(yàn)總結(jié)（2）傅里葉形狀描述原理傅里葉形狀描述符(Fourier shape descriptors)基本思想是用物體邊界的傅里葉變換作為形狀描述，利用區(qū)域邊界的封閉性和周期性，將二維問題轉(zhuǎn)化為一維問題。由邊界點(diǎn)導(dǎo)出三種形狀表達(dá)，分別是曲率函數(shù)、質(zhì)心距離、復(fù)坐標(biāo)函數(shù)。（3）幾何參數(shù)原理形狀的表達(dá)和匹配采用更為簡(jiǎn)單的區(qū)域特征描述方法，例如采用有關(guān)形狀定量測(cè)度（如矩、面積、周

32、長(zhǎng)等）的形狀參數(shù)法（shape factor）。在 QBIC 系統(tǒng)中，便是利用圓度、偏心率、主軸方向和代數(shù)不變矩等幾何參數(shù)，進(jìn)行基于形狀特征的圖像檢索。實(shí)驗(yàn)五 攝像機(jī)標(biāo)定實(shí)驗(yàn)內(nèi)容輔導(dǎo)材料一、實(shí)驗(yàn)?zāi)康?#160; 1、熟悉并掌握MATLAB圖像處理工具箱的使用；2、理解攝像機(jī)標(biāo)定（DLT）方法的原理，并編程實(shí)現(xiàn)攝像機(jī)內(nèi)參數(shù)和外參數(shù)的估計(jì)。二、實(shí)驗(yàn)環(huán)境MATLAB R2008以上版本、WIN7計(jì)算機(jī)三、實(shí)驗(yàn)內(nèi)容 1)讀取一幅圖像并顯示。myread=imread('a1.jpg');imshow(myread)2)檢查內(nèi)存（數(shù)組）中的圖像。>> whos Name Si

33、ze Bytes Class Attributes myread 768x1024x3 2359296 uint8 3)讀取圖像中像素點(diǎn)的坐標(biāo)值。clcclearimshow('a1.jpg')hold onx_total=;y_total=;a=input('輸入要讀取的像素點(diǎn)個(gè)數(shù)，并選取(1-100): ')for i=1:ax,y=ginput(1);plot(x,y,'+r')text(x+0.5,y+0.5,num2str(i)x_total=x_total,x;y_total=y_total,y;enddisp('x='

34、;)disp(x_total)disp('y=')disp(y_total)以取6個(gè)點(diǎn)為例，如下圖所示：各點(diǎn)坐標(biāo)為：x= 538.0000 613.0000 682.0000 391.0000 473.5000 266.5000y= 718.2500 656.7500 593.7500 650.7500 487.2500 293.75004)獲取標(biāo)定塊圖像的64個(gè)特征點(diǎn)坐標(biāo)。clcclearimshow('a3.jpg')hold onx_total=;y_total=;a=input('請(qǐng)輸入標(biāo)定點(diǎn)個(gè)數(shù)并標(biāo)定(1-100): ')for i=1

35、:ax,y=ginput(1);plot(x,y,'or')text(x+0.5,y+0.5,num2str(i)x_total=x_total,x;y_total=y_total,y;enddisp('x=')disp(x_total)disp('y=')disp(y_total)可得到每個(gè)點(diǎn)對(duì)應(yīng)的圖像坐標(biāo)如下：x= Columns 1 through 9 538.0000 577.0000 613.0000 649.0000 682.0000 718.0000 748.0000 778.0000 535.0000 Columns 10 thr

36、ough 18 484.0000 434.5000 388.0000 341.5000 298.0000 254.5000 214.0000 532.0000 574.0000 Columns 19 through 27 611.5000 649.0000 683.5000 718.0000 749.5000 782.5000 532.0000 478.0000 428.5000 Columns 28 through 36 379.0000 331.0000 286.0000 242.5000 200.5000 527.5000 571.0000 610.0000 647.5000 Colum

37、ns 37 through 45 683.5000 718.0000 754.0000 785.5000 527.5000 470.5000 419.5000 368.5000 322.0000 Columns 46 through 54 274.0000 227.5000 184.0000 523.0000 566.5000 607.0000 646.0000 683.5000 721.0000 Columns 55 through 63 755.5000 790.0000 521.5000 464.5000 412.0000 359.5000 310.0000 260.5000 215.5

38、000 Column 64 173.5000y= Columns 1 through 9 718.2500 688.2500 655.2500 623.7500 596.7500 566.7500 539.7500 515.7500 668.7500 Columns 10 through 18 646.2500 623.7500 601.2500 580.2500 559.2500 541.2500 518.7500 616.2500 584.7500 Columns 19 through 27 553.2500 524.7500 496.2500 469.2500 442.2500 419.

39、7500 560.7500 541.2500 518.7500 Columns 28 through 36 497.7500 476.7500 458.7500 439.2500 421.2500 503.7500 475.2500 445.2500 418.2500 Columns 37 through 45 389.7500 364.2500 338.7500 317.7500 448.2500 427.2500 406.2500 388.2500 368.7500 Columns 46 through 54 349.2500 332.7500 316.2500 388.2500 359.

40、7500 332.7500 305.7500 278.7500 254.7500 Columns 55 through 63 230.7500 209.7500 323.7500 307.2500 287.7500 271.2500 253.2500 236.7500 220.2500 Column 64 205.25005)編程估計(jì)攝像機(jī)內(nèi)參數(shù)和外參數(shù)。首先確定11小題中的特征點(diǎn)在世界坐標(biāo)系中的坐標(biāo)：設(shè)每小格的長(zhǎng)度為10建立如下圖所示坐標(biāo)是：由此可以得到64個(gè)點(diǎn)的世界坐標(biāo)是為：world_x=0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7.

41、 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0;world_y=0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7. 0 0 0 0 0 0 0 0. 0 1 2 3 4 5 6 7;world_z=0 0 0 0 0 0 0 0. 1 1 1 1 1 1 1 1. 2 2 2 2 2 2 2 2. 3 3 3 3 3 3 3

42、 3. 4 4 4 4 4 4 4 4. 5 5 5 5 5 5 5 5. 6 6 6 6 6 6 6 6. 7 7 7 7 7 7 7 7;程序：x = 538.0000 577.0000 613.0000 649.0000 682.0000 718.0000 748.0000 778.0000 535.0000. 484.0000 434.5000 388.0000 341.5000 298.0000 254.5000 214.0000 532.0000 574.0000. 611.5000 649.0000 683.5000 718.0000 749.5000 782.5000 532.

43、0000 478.0000 428.5000. 379.0000 331.0000 286.0000 242.5000 200.5000 527.5000 571.0000 610.0000 647.5000. 683.5000 718.0000 754.0000 785.5000 527.5000 470.5000 419.5000 368.5000 322.0000. 274.0000 227.5000 184.0000 523.0000 566.5000 607.0000 646.0000 683.5000 721.0000. 755.5000 790.0000 521.5000 464.5000 412.0000 359.5000 310.0000 260.5000 215.5000. 173.5000;y=718.2500 688.2500 655.2500 623.7500 596.7500 566.7500 539.7500 515.7500 668.7500. 646.2500 623.7500 601.2500 580.2500 559.2500 541.2500 518.7500 616.2500 584.7500. 553.2500 524.7500 496.2500 469.2500 442.2