Array indices must be positive integers or logical values.

Hello everyone

Ive been trying to make a project that recognize letters from handwriting

although i am really stuck because of an error that has caused

i am using neural network that i am training with data base from the internet

but then i am trying to separate each letter that has been recognized and feed it to the neural network but i am getting an error while doing that

i would be glad if someone could help me with this one . i have seen other posts with same problem but couldnt understand the solution

thanks in adavance

the error:

Array indices must be positive integers or logical values.

Error in Network (line 106)
    outputMatrix=net(segImage);
 

the code :

    
close all
clear
clc





digitDatasetPath = fullfile('C:\Users\user\Downloads\MatlabCentralUpload (1)\MatlabCentralUpload\HandWriting Recognition Project\archive (1)');
imds = imageDatastore(digitDatasetPath,'IncludeSubfolders',true,'LabelSource','foldernames');
labelCount = countEachLabel(imds);
img = readimage(imds,1);
size(img);
numTrainFiles = 300;
[imdsTrain,imdsValidation] = splitEachLabel(imds,numTrainFiles,'randomize');

layers = [
    imageInputLayer([34 34 1])
    
    convolution2dLayer(3,8,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(3,16,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(3,32,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    fullyConnectedLayer(26)
    softmaxLayer
    classificationLayer];

options = trainingOptions('sgdm', ...
    'InitialLearnRate',0.01, ...
    'MaxEpochs',4, ...
    'Shuffle','every-epoch', ...
    'ValidationData',imdsValidation, ...
    'ValidationFrequency',30, ...
    'Verbose',false, ...
    'Plots','training-progress');

net = trainNetwork(imdsTrain,layers,options);

YPred = classify(net,imdsValidation);
YValidation = imdsValidation.Labels;

accuracy = sum(YPred == YValidation)/numel(YValidation);


%%Read Image
imagen=imread('patna.png');
%%space
sp=10;
%%Show image
figure(1)
imshow(imagen);
title('INPUT IMAGE WITH NOISE')
%%Convert to gray scale
if size(imagen,3)==3 % RGB image
    imagen=rgb2gray(imagen);
end
%%Convert to binary image
threshold = graythresh(imagen);
imagen =~im2bw(imagen,threshold);
%imagen = imclose(imagen, strel('rectangle',[3 ceil(Sp/2)]));
%%Remove all object containing fewer than 30 pixels
imagen = bwareaopen(imagen,15);
pause(1)
%%Show image binary image
figure(2)
imshow(~imagen);
title('INPUT IMAGE WITHOUT NOISE')
%%Label connected components
[L Ne]=bwlabel(imagen);
%%Measure properties of image regions
propied=regionprops(L,'BoundingBox');
hold on
%%Plot Bounding Box
for n=1:size(propied,1)
    rectangle('Position',propied(n).BoundingBox,'EdgeColor','g','LineWidth',2)
end
hold off
pause (1)
%%Objects extraction
figure
for n=1:Ne
    [r,c] = find(L==n);
    n1=imagen(min(r):max(r),min(c):max(c));
    imshow(~n1);
    fullFileName = fullfile('segmentedImages', sprintf('image%d.png', n));
    imwrite(n1, fullFileName);
    pause(0.5)
end

%% Feeding to Neural Network and Detecting Text
for i=1:Ne
    segImage=double(imread(fullfile('segmentedImages', sprintf('image%d.png', i))));
    outputMatrix=net(segImage);
    row=find(ismember(outputMatrix, max(outputMatrix(:)))); % returns the row number which has highest probability
%     figure(i);
%     colormap(gray)
    character = double(imread(fullfile('segmentedImages', sprintf('image%d.png', i))));
%     imagesc(~character)
%     title(imageLabeler(row))
    detectedWord(1,i)=imageLabeler(row);
end

Answer 1

[yanqi liu]

yes，sir，if use this cnn net to classify，may be use

[YPred,scores] = classify(net,segImage)

[tamir bonavida]

Hello Thank you for your help

i did add the line you recommended but now i have another Error

Error using categorical/max (line 87)
Relational comparisons are not allowed for categorical arrays that are not ordinal.
Error in Network (line 109)
   row=find(ismember(YPred, max(YPred(:)))); % returns the row number which has highest probability
   
   
   
   
%% Feeding to Neural Network and Detecting Text
for i=1:Ne
    segImage=double(imread(fullfile('segmentedImages', sprintf('image%d.png', i))));
    [YPred,scores] = classify(net,segImage);
    row=find(ismember(YPred, max(YPred(:)))); % returns the row number which has highest probability
%     figure(i);
%     colormap(gray)
    character = double(imread(fullfile('segmentedImages', sprintf('image%d.png', i))));
%     imagesc(~character)
%     title(imageLabeler(row))
    detectedWord(1,i)=imageLabeler(row);
end

[yanqi liu]

yes，sir，if possible，may be upload your data to debug

when use classify，the return value is its class result，so may be not need use find(ismember) to get the index row

[tamir bonavida]

the data base was downloaded from here

when i try to extract the classification I only get the letter 'L'

do you think there is a problem with the net ?

i extracted the classification like this

for i=1:Ne
    segImage=double(imread(fullfile('segmentedImages', sprintf('image%d.png', i))));
    [YPred,scores] = classify(net,segImage);
    disp(title(char(YPred)));

end