Health & Fitness

Optical Character Recognition using Template Matching & Structured Analysis

Description
Optical character recognition (OCR) is an important research area in pattern recognition. The objective of an OCR system is to recognize alphabetic letters, numbers, or other characters, which are in the form of digital images, without any human
Published
of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Optical Character Recognition using Template Matching &Structured Analysis Desai Ami S.Assi. Prof. in Vie!anand college for "CA#Surat ami$aidya%gmail.com A"STRACT Optical character recognition (OCR) is an important research area in pattern recognition.The objective of an OCR system is to recognize alphabetic letters, numbers, or other characters, which are in the form of igital images, without any human intervention. Thisis accomplishe by searching a match between the features e!tracte from the givencharacter"s image an the library of image moels. #eally, we woul li$e the features to be istinct for ifferent character images so that the computer can e!tract the correctmoel from the library without any confusion. %t the same time, we also want thefeatures to be robust enough so that they will not be affecte by viewing transformations,noises, resolution variations an other factors. &igure illustrates the basic processes of anOCR system. The asic process of an OCR system 'ifferent &amilies of Character Recognition  &igure shows the ifferent families of character recognition. Two ifferentfamilies are inclue in the general term of character recognition  On*line character recognition  Off*line character recognition • T'MP(AT' MATC)*+, Optical Character Recognition by using template matching is a system prototype that is useful to recognize the character or alphabet by comparing + images of the alphabet. The objectives of this system prototype are to evelop a prototype for the Optical Character Recognition (OCR) system an to implement the Template atching as the algorithm that applie to recognize the characters- characters to be teste are alphabet (%*), grey / scale images. The purpose of this system prototype is to solve the problem in recognizingthe character which is ifficult to recognize before. Objective i) To evelop a prototype of Optical Character Recognition (OCR) system. ii) To apply a Template atching approach in recognizing character.0cope i) Template atching is the algorithm that applie to recognize the characters. ii) Characters to be teste are alphabet (% * ). iii) 1rey*scale images were use with Times 2ew Roman font type. iv) 3sing bitmap image format v) 3sing +45 ! +45 image size 6OR7&8O6 O& T9: T:;8%T: %TC9#21%81OR#T  -igure 6or$flow of the Template atching %lgorithm T'MP(AT' MATC)*+, ST'PS -OR R'CO,+*T*O+ The template*matching algorithm implements the following steps i) &irstly, the character image from the etecte string is selecte. ii) %fter that, the image to the size of the first template is rescale. iii) %fter rescale the image to the size of the first template (srcinal)image, the matching metric is compute. iv) Then the highest match foun is store. #f the image is not match  repeat again the thir step. v) The ine! of the best match is store as the recognize character.   The value of the ata that was entere be e!tracte from the images, comprising8etters. :ach character was automatically selecte an threshol using methos  previously escribe. :!traction of the image of the character. The image is converte.<itmap is represente by =+!=+*matri! or by =44 vectors with 5 an = .   This process involves the use of a atabase of characters or templates. Theree!ists a template for all possible input characters. &or recognition to occur, the currentinput character is compare to each template to fin either an e!act match, or thetemplate with the closest representation of the input character. #f  I(x, y)  is the inputcharacter, Tn(x, y)  is the template n, then the matching function  s(I, Tn)   will return avalue inicating how well template n  matches the input character .0ome of the morecommon matching functions are base on the following formulas   Character recognition is achieve by ientifying which Tn  gives the best value of matching function,  s(I ,Tn)  . The metho can only be successful if the input character can the store templates are of the same or similar font. Template matching can be performe on binary, threshol characters or on gray*level characters. &or gray*level characters, it is more common for 2ormalize Correlation to be use as this. -*(T'R*+, *MA,'   Some of the algorithm that was applied in the technique issuch as below: Aerage filtering techni/ue algorithm )>(),( = )5,( +  ∑ = W r  cr d  N S  ε    The secon process after filtering the image is the threshol techni?ue. There alot of techni?ues for threshol the image such as the minimum threshol, ma!imumthreshol, meian threshol an the average threshol. Threshol techni?ue is the techni?ue that use to transform a grayscale imageinto a binary image which is using (= or 5) values. % threshol is set which each pi!el iscompare to other pi!els. #f the pi!el is greater than or e?ual to this threshol, it isoutputte as a =. Otherwise it is outputte as a 5. Threshol converts each pi!el into blac$, white or unchange epening on whether the srcinal color value is within thethreshol range.
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks