movistar_captcha

readme.md

Simple Captcha

This uses ImageMagick and Tesseract to break Movistar's Captcha from their SMS system.

Command Line Process

File movistar.sh.

Raw Image

Clean Image

After running convert in.png -fill white -opaque '#c0c0c0' in.tiff

OCR

After running tesseract in.tiff text we get the text in the file result.txt

$ cat result.txt
IKEUQ

Ruby Script

Check the file movistar.rb. This generates a log file log.html

movistar.rb

# Jaime Moran
require 'net/http'

def download_captcha file_name
  uri      = URI.parse('http://www.corporativo.telefonica.com.sv/EnviarSMSSV/faces/' +
                      'EnviarSMS.jsp?vrClientId=form1:imageEx1')
  headers  = { 'user-agent' =>  'Mozilla/4.0 (Compatible; MSIE 4.0)' }
  http     = Net::HTTP.start(uri.host, uri.port)
  response = http.request_get(uri.path, headers)
  
  # set cookie in the header to be used by the new request
  headers['cookie'] = response['set-cookie']
  
  # request captcha image
  resimg = http.request_get( uri.path + '?' + uri.query, headers) 

  # store the image on disk
  open( file_name, 'w'){ |file| file.write(resimg.body) }

end

# clean and OCR image.
def process_image file_name
  temp_file = '_temp.tiff'
  temp_set  = '_result'

  # convert image to tiff, remove gray color and store text into set
  system [ 'convert', file_name, '-fill white -opaque "#c0c0c0"', temp_file, 
           '&& tesseract', temp_file, temp_set ].join ' '

  text  =  IO.read "#{temp_set}.txt"

  # clean up
  system "rm #{temp_file} #{temp_set}.txt"

  text

end

# append to log
def log file_name, text
  open( "log.html", 'a+' ){ |file|
    file.write("<p> <img src='#{file_name}'/>#{text.strip}</p> \n")
  }
end

# #Running
file_name = "#{Time.now.to_f}.png"
download_captcha(file_name)
text = process_image(file_name)
log(file_name, text)

movistar.sh

convert in.png -fill white -opaque '#c0c0c0' in.tiff && tesseract in.tiff text && cat text.txt