Problem With Vehicle number plate detection code.

Hello, so im trying to use a vehicle number plate detection from Pankaj Katri, but there are errors to when i run it.

Here are the errors that showed and cant understand.

(this error came from running number 3 which is Plate_detection.m which the authors says to use to run the code)

Plate_detection
Brace indexing is not supported for variables of this type.

Error in Letter_detection (line 8)
cor=corr2(NewTemplates{1,n},snap);

Error in Plate_detection (line 34)
letter=Letter_detection(Iprops(i).Image); % Reading the letter corresponding the binary image 'N'.

you can easily find the website by searching "Car Number Plate Detection using MATLAB and Image Processing"

i am currently using the latest matlab which is R2021b

(the Zip is now included in this post)

Here the the three matlab codes that the author made

1. template_creation.m

%CREATE TEMPLATES 
%Alphabets
A=imread('alpha/A.bmp');B=imread('alpha/B.bmp');C=imread('alpha/C.bmp');
D=imread('alpha/D.bmp');E=imread('alpha/E.bmp');F=imread('alpha/F.bmp');
G=imread('alpha/G.bmp');H=imread('alpha/H.bmp');I=imread('alpha/I.bmp');
J=imread('alpha/J.bmp');K=imread('alpha/K.bmp');L=imread('alpha/L.bmp');
M=imread('alpha/M.bmp');N=imread('alpha/N.bmp');O=imread('alpha/O.bmp');
P=imread('alpha/P.bmp');Q=imread('alpha/Q.bmp');R=imread('alpha/R.bmp');
S=imread('alpha/S.bmp');T=imread('alpha/T.bmp');U=imread('alpha/U.bmp');
V=imread('alpha/V.bmp');W=imread('alpha/W.bmp');X=imread('alpha/X.bmp');
Y=imread('alpha/Y.bmp');Z=imread('alpha/Z.bmp');

%Natural Numbers
one=imread('alpha/1.bmp');two=imread('alpha/2.bmp');
three=imread('alpha/3.bmp');four=imread('alpha/4.bmp');
five=imread('alpha/5.bmp'); six=imread('alpha/6.bmp');
seven=imread('alpha/7.bmp');eight=imread('alpha/8.bmp');
nine=imread('alpha/9.bmp'); zero=imread('alpha/0.bmp');

%Creating Array for Alphabets
letter=[A B C D E F G H I J K L M N O P Q R S T U V W X Y Z];
%Creating Array for Numbers
number=[one two three four five six seven eight nine zero];

NewTemplates=[letter number];
save ('NewTemplates','NewTemplates')
clear all

2. Letter_detection.m

function letter=readLetter(snap)

load NewTemplates 
snap = imresize(snap,[42 24]); 
rec=[ ];

for n=1:length(NewTemplates)
    cor=corr2(NewTemplates{1,n},snap); 
    rec=[rec cor]; 
end

ind=find(rec==max(rec));
display(ind);

% Alphabets listings.
if ind==1 || ind==2
    letter='A';
elseif ind==3 || ind==4
    letter='B';
elseif ind==5
    letter='C';
elseif ind==6 || ind==7
    letter='D';
elseif ind==8
    letter='E';
elseif ind==9
    letter='F';
elseif ind==10
    letter='G';
elseif ind==11
    letter='H';
elseif ind==12
    letter='I';
elseif ind==13
    letter='J';
elseif ind==14
    letter='K';
elseif ind==15
    letter='L';
elseif ind==16
    letter='M';
elseif ind==17
    letter='N';
elseif ind==18 || ind==19
    letter='O';
elseif ind==20 || ind==21
    letter='P';
elseif ind==22 || ind==23
    letter='Q';
elseif ind==24 || ind==25
    letter='R';
elseif ind==26
    letter='S';
elseif ind==27
    letter='T';
elseif ind==28
    letter='U';
elseif ind==29
    letter='V';
elseif ind==30
    letter='W';
elseif ind==31
    letter='X';
elseif ind==32
    letter='Y';
elseif ind==33
    letter='Z';
    %*-*-*-*-*
% Numerals listings.
elseif ind==34
    letter='1';
elseif ind==35
    letter='2';
elseif ind==36
    letter='3';
elseif ind==37 || ind==38
    letter='4';
elseif ind==39
    letter='5';
elseif ind==40 || ind==41 || ind==42
    letter='6';
elseif ind==43
    letter='7';
elseif ind==44 || ind==45
    letter='8';
elseif ind==46 || ind==47 || ind==48
    letter='9';
else
    letter='0';
end
end

3. Plate_detection.m

im1 = imread('BluePN.JPG');
imgray = rgb2gray(im1);
imbin = im2bw(imgray);
im = edge(imgray, 'prewitt');

%Below steps are to find location of number plate
Iprops1=regionprops(im,'BoundingBox','Area', 'Image');
area1 = Iprops1.Area;
count1 = numel(Iprops1);
maxa= area1;
boundingBox1 = Iprops1.BoundingBox;
for i=1:count1
   if maxa<Iprops1(i).Area
       maxa=Iprops1(i).Area;
       boundingBox1=Iprops1(i).BoundingBox;
   end
end    

im = imcrop(imbin, boundingBox1);%crop the number plate area
im = bwareaopen(~im, 500); %remove some object if it width is too long or too small than 500

 [h, w] = size(im);%get width

imshow(im);

Iprops=regionprops(im,'BoundingBox','Area', 'Image'); %read letter
count = numel(Iprops);
noPlate=[]; % Initializing the variable of number plate string.

for i=1:count
   ow = length(Iprops(i).Image(1,:));
   oh = length(Iprops(i).Image(:,1));
   if ow<(h/2) && oh>(h/3)
       letter=Letter_detection(Iprops(i).Image); % Reading the letter corresponding the binary image 'N'.
       noPlate=[noPlate letter]; % Appending every subsequent character in noPlate variable.
   end
end

Thank you for taking the time to read this.

[KSSV]

We cannot help you unless whole code with the inputs is known to us.

[Matt Alvarez]

ah im sorry will edit the post for complete information thank you for your comment!

[yanqi liu]

yes，sir，may be add the m files and images to zip file，then upload the zip file，so we make some debug

[Matt Alvarez]

ah yes i will edit the post again to add the zip file. Thank you for your comment!

Answer 1

[Harikrishnan Balachandran Nair]

Hi Matt,

In the code that you have provided, the variable 'NewTemplates' is being stored as an array of images, which are of the same size. But, for calculating the 2D correlation coefficient , brace indexing is being used to access the array elements. This gives rise to error as NewTemplates is not stored as a 'cell array' . You can change it to the following line of code to avoid the error :

cor=corr2(NewTemplates(1,n),snap); 

Hope this helps!