Undefined function or variable 'A'.

clc;
clear all;
close all
%------------Image Reading------------
[FILENAME,PATHNAME]=uigetfile('*.jpg','Select the Image');
FilePath=strcat(PATHNAME,FILENAME);
disp('The Image File Location is');
disp(FilePath);
[DataArray,map]=imresize(imread(FilePath),[300,650]);
figure,imshow(DataArray,map);
title('Input Image');
% Separate Channel
r_channel=DataArray(:,:,1);
b_channel=DataArray(:,:,2);
g_channel=DataArray(:,:,3);
% Noise Remaoval
r_channel=medfilt2(r_channel);
g_channel=medfilt2(g_channel);
b_channel=medfilt2(b_channel);
% restore channels
rgbim(:,:,1)=r_channel;
rgbim(:,:,2)=g_channel;
rgbim(:,:,3)=b_channel;
figure,imshow(uint8(rgbim));
title('Denoised Image');
% RGB to Gray
Igray = 0.30*r_channel + 0.59*g_channel + 0.11*b_channel;
figure,imshow(uint8(Igray));
title('Gray Image');
% Edge Detection
y=double(Igray);
f1 = zeros(3,3,5);
f1(:,:,1) = [1 2 1;0 0 0;-1 -2 -1]; %vertical
f1(:,:,2) = [-1 0 1;-2 0 2;-1 0 1];  %horizontal
f1(:,:,3) = [2 2 -1;2 -1 -1;-1 -1 -1]; % 45 diagonal
f1(:,:,4) = [-1 2 2; -1 -1 2;-1 -1 -1]; %135 diagonal
f1(:,:,5) = [-1 0 1;0 0 0;1 0 -1]; % non directional
for i = 1:5
    g_im(:,:,i) = filter2(f1(:,:,i),y);
end
[m, p] = max(g_im,[],3);
edim = edge(y, 'canny');
im2 = (p.*edim);
edhist=im2;
figure, imshow(edhist)
title('Edge Detection');
Avg=mean2(Igray);
if(Avg>202 && Avg<207)
    load 100.mat
elseif(Avg>175 && Avg<180)
    load 200.mat
elseif(Avg>190 && Avg<195)
    load 500.mat
elseif(Avg>209 && Avg<214)
    load 2000.mat
end
figure,imshow(uint8(DataArray));
title('ROI-Extract Texture & Statistical Features');
hold on
for n=1:size(A,1)
    rectangle('Position',A(n,:),'EdgeColor','r','LineWidth',4)
end
pause(1)
figure,imshow(edhist);
title('ROI-Extract Edge & Shape Features');
hold on
for n=1:size(A,1)
    rectangle('Position',A(n,:),'EdgeColor','r','LineWidth',4)
end
pause(1)
SFL_Data=zeros(size(A,1),6);
SSL_Data=zeros(size(A,1),12);
for n=1:size(A,1)
    imcropgray = imcrop(Igray,A(n,:));
    Img_data=imcropgray;
    % STATISTICAL FEATURES
    % First Level Feature
    Mean = mean2(Img_data);
    Variance  = mean2(var(double(Img_data)));
    Kurtosis = kurtosis(double(Img_data(:)));
    stats = graycoprops(Img_data,'Contrast Correlation Energy Homogeneity');
    Energy = stats.Energy;
    Contrast = stats.Contrast;
    Entropy = entropy(Img_data);
    FL_Feat=[Mean Variance Kurtosis Energy Contrast Entropy];
    FL_Feat(isnan(FL_Feat))=0;
    %   disp('First Level Feature");
    %   disp(FL_Feat)
    SFL_Data(n,:)=FL_Feat;
    %Second Level Feature
    offsets = [0 1; -1 1; -1 0; -1 -1];  %0°, 45°, 90°, 135°,  
    GLCM1 = graycomatrix(Img_data,'NumLevels',8,'Offset',offsets);
    GLCM2 = graycomatrix(Img_data,'NumLevels',32,'Offset',offsets);
    stats = graycoprops(GLCM1,'Contrast Correlation Energy Homogeneity');
    stats1 = graycoprops(GLCM2,'Contrast Correlation Energy Homogeneity');
    Correlation=[mean(stats.Correlation) mean(stats1.Correlation)];
    ASM=[mean(stats.Energy) mean(stats1.Energy)];
    Homogeneity=[mean(stats.Homogeneity) mean(stats1.Homogeneity)];
    IDM=[Inverse_Diff(GLCM1) Inverse_Diff(GLCM2)];
    Max_prob=[Maximum_Prob(GLCM1) Maximum_Prob(GLCM2)];
    Entropy = [entropy(GLCM1) entropy(GLCM2)];
    SL_Feat=[ASM Correlation Homogeneity IDM Max prob Entropy];
    SL_Feat(isnan(SL_Feat))=0;
    %     disp('Second Level Feature')
    %     disp(SL_Feat)
    SSL_Data(n,:)=SL_Feat;
end
ST_feat=[mean(SFL_Data) mean(SSL_Data)];
disp('Statistical Features')
disp(ST_feat);
EF_Data=zeros(size(Avg,l),7);
for n=l:size(Avg,l) 
    imcropedge=imcrop(edhist,Avg(n,:));
    % Edge Features
    results=regionprops(imcropedge,'Area','EulerNumber','Orientation','BoundingBox','Extent',...
        'Perimeter','Centeroid','Extrema','PixelIdxList','ConvexArea',...
        'FilledArea','PixelList','ConvexHull','FilledImage','Solidity',...
        'ConvexImage','Image','SubarrayIdx','Eccentricity','MajorAxisLength',...
        'EquivDiameter','MinorAxisLength','EulerNumber');
    NR=vertcat(results.BoundingBox);
    Circularity=zeros(size(NR,1));
    Eccentricity=zeros(size(NR,1));
    Convexity=zeros(size(NR,1));
    Area=zeros(size(NR,1));
    Compactness=zeros(size(NR,1));
    Extent=zeros(size(NR,1));
    Solidity=zeros(size(NR,1));
    for ii=1:size(NR,1)
        Circularity(ii) = ((results(ii).Perimeter).^2 )./ (4* (pi*(results(ii).Area)));
        Circularity(isnan(Circularity))=0;
        Circularity(isinf(Circularity)) = 0;
        Compactness(ii)=(4*results(ii).Area*pi)/(results(ii).Perimeter).^2;
        Compactness(isnan(Compactness))=0;
        Compactness(isinf(Compactness)) = 0;
        Convexity(ii)=results(ii).ConvexArea;
        Convexity(isnan(Convexity))=0;
        Convexity(isinf(Convexity)) = 0;
        Area(ii)=results(ii).Area;
        Area(isnan(Area))=0;
        Area(isinf(Area)) = 0;
        Eccentricity(ii)=results(ii).Eccentricity;
        Eccentricity(isnan(Eccentricity))=0;
        Eccentricity(isinf(Eccentricity)) = 0;
        Extent(ii)=results(ii).Extent;
        Extent(isnan(Extent))=0;
        Extent(isinf(Extent)) = 0;
        Solidity(ii)=results(ii).Solidity;
        Solidity(isnan(Solidity))=0;
        Solidity(isinf(Solidity)) = 0;
    end
    SF=[mean2(Area) mean2(Solidity) mean2(Convexity) mean2(Circularity) mean2(Eccentricity) mean2(Compactness) mean2(Extent)];
    EF_Data(n,:)=SF;
end
EDF_feat=mean(EF_Data);
disp('Edge Features')
Tfeat=[ST_feat EDF_feat];
load Pdata.mat
load Ndata.mat
xdata = [Train_dataP;Train_dataN];
group = [Train_LabP;Train_LabN];
fitcsvmTrain = fitcsvmtrain(xdata,group,'kernel_function','rbf');
Classify_Result= fitcsvmclassify(svmTrain,Tfeat);
if(Classify_Result==1)
    figure,imshow(DataArray,map);
    title('Currency Type: Real');
    msgbox('Currency Type: Real');
else
    figure,imshow(DataArray,map);
    title('Currency Type: Fake');
    msgbox('Currency Type: Fake');
end
% pause(2)
% perfdata

Answer 1

[Walter Roberson]

Your code appears to count on there being a variable named A loaded from 100.mat or 200.mat or 500.mat or 2000.mat (depending on what value was calculated), but whichever .mat file was loaded did not contain that variable.

whos -file 100.mat
whos -file 200.mat
whos -file 500.mat
whos -fiel 2000.mat

We recommend that you do not use the command form of load of .mat files, that you use the function form, such as

datacell = load('100.mat');
assert(isfield(datacell, 'A'), 'missing variable A in mat file');
A = datacell.A;

Answer 2

[Torsten]

Seems your MATLAB files *.mat don't include a matrix "A". Or the value of "Avg" is out of range.

if(Avg>202 && Avg<207)
    load 100.mat
elseif(Avg>175 && Avg<180)
    load 200.mat
elseif(Avg>190 && Avg<195)
    load 500.mat
elseif(Avg>209 && Avg<214)
    load 2000.mat
end