When I looked at their (ifnenit) website:

http://www.ifnenit.com/ground_truth.htm

I can see that you first need to segment the characters then recognize them and finally compare them to the ground truth (#2 in the image below).

In the image:

#1 Characters

#2 the whole word

#3 The base-line coordinates I suppose

#4 number of characters

@Abbas Cheddad If I want to extract number of characters (#4) or recognize them (#1) automatically from this file,WHAT SHALL I DO.

Extracting number of characters (code in MATLAB):

fileID = fopen('ae07_001.tru','r');

formatSpec = '%s';

text = fscanf(fileID,formatSpec);

fclose(fileID);

%The # is between CHA and BLN

ind=strfind(text ,'CHA');

ind2=strfind(text ,'BLN');

numOfCharacters = str2double(text((ind+4):(ind2-1)));

Thank you for the valuable information and quick response.

I am working by python, If you want I can share code with you to read the ground truth

@ Mohammed Aweny, yes share me python code

from pathlib import Path

directiory_a = Path('Your Path/set_a/tru')

directiory_b = Path('Your Path/set_b/tru')

directiory_c = Path('Your Path/set_c/tru')

directiory_d = Path('Your Path/set_d/tru')

directiory_e = Path('Your Path/set_e/tru')

files_a = os.listdir(directiory_a) # list of all file names in directiory_a

files_b = os.listdir(directiory_b)

files_c = os.listdir(directiory_c)

files_d = os.listdir(directiory_d)

files_e = os.listdir(directiory_e)

zip_code_a = []

zip_code_b = []

zip_code_c = []

zip_code_d = []

zip_code_e = []

for file in files_a:

with open(directiory_a/file, 'r') as F:

lines = F.readlines()

line = lines[6]

[zip_code_a.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()]

print(len(zip_code_a))

for file in files_b:

with open(directiory_b/file, 'r') as F:

lines = F.readlines()

line = lines[6]

[zip_code_b.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()]

print(len(zip_code_b))

for file in files_c:

with open(directiory_c/file, 'r') as F:

lines = F.readlines()

line = lines[6]

[zip_code_c.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()]

print(len(zip_code_c))

for file in files_d:

with open(directiory_d/file, 'r') as F:

lines = F.readlines()

line = lines[6]

[zip_code_d.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()]

print(len(zip_code_d))

for file in files_e:

with open(directiory_e/file, 'r') as F:

lines = F.readlines()

line = lines[6]

[zip_code_e.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()]

print(len(zip_code_e))

@ Nisreen AbdAllah Ali

Mohamed Awni you missed "import os" in the header of file. I have just run the code and the output was numbers from 1 to 569, number of files in the tru folder. What shall I change to read line 6 which contain [ LBL: ZIP:6056;AW1: ] and the only four digits between (:) and (;). Please I need your help.

Where can I find set_b, set_c, set_d, and set_e of the data-set.

Is it free?! from where I can download them??

Thank you in advance.

You have the demo version of this dataset, you can send an E-mail and register through this link " http://www.ifnenit.com/register.htm ".

You can read the line 6 and extract the zip code by

"

line = lines[6] # read the line no. 6 from all lines in txt file

[zip_code_e.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()] # extract the 4 digits zip code@ Nisreen AbdAllah Ali

"

Mohamed Awni output of the above code is not four digits. It gives me the sequence numbers from 1 to 569. I removed the (len) from last line "print(len(zip_code_a))" but the program hang.

what is the output of the code after run Module?

It is not match the four digits for ZIP in the line (LBL)?

This is an example of line no. 6 "LBL: ZIP:1251;AW1:" right?

and this is the code " [zip_code_e.append(int(s)) for s in line.split(":")[2].split(';') if s.isdigit()] "

line >>> LBL: ZIP:1251;AW1:

split it by ":" give alist of [LBL, ZIP, 1251;, AW1] #list contains 4 items this list of position 2 gives 1251; which need to split it again by ";" gives 1251 then convert to to int and append it to list "zip_code_e"@ Nisreen AbdAllah Ali