Research Cell: An International Journal of Engineering Sciences

Bharti Mehta ¹, Simpel Rani ²

Affiliations
1 Department of Computer Engineering, Yadavindra College of Engineering, Talwandi Sabo (Bathinda), IN
2 Department of Computer Engineering, Yadavindra College of Engineering, Talwandi Sabo (Bathinda), IN

Abstract

Character Segmentation of Handwritten Documents has been an active area of research and due to its diverse applicable environment; it continues to be a challenging research topic. The desire to edit scanned text document forces the researchers to think about the optical character recognition (OCR). OCR is the process of recognizing a segmented part of the scanned image as a character. OCR process consists of three major sub processes - pre processing, segmentation and then recognition. Out of these three, the segmentation process is the most important phase of the overall OCR process. Different problems in the characters segmentation of handwritten text is due to the different writing style of different people because the size and shape is not fixed while we write any text. In this work, we formulate an algorithm to segment the scanned document image as a character. According to proposed algorithm, broken characters in Gurmukhi script, we used the segmentation of these characters that can become easily identify how many characters are in one word. To develop the algorithm to segment the characters from a word we are using combinations of two approaches which are Horizontal Profile Projection and Vertical Profile Projection. And get the accuracy is 93%.

Keywords

Gurmukhi Script, OCR, Segmentation, Handwritten Document, Horizontal Profile Projection, Vertical Profile Projection.

Full Text

   

Neetika Bansal ¹, Vishal Goyal ², Simpel Rani ³

Affiliations
1 Department of Computer Science, Punjabi University College of Engineering & Management, Rampura Phul, IN
2 Department of Computer Science Punjabi University, Patiala, IN
3 Department of Computer Science and Engineering, Yadavindra College of Engineering, Talwandi Sabo, IN

Abstract

Code mixing denotes using multiple languages in an utterance. It is clearly seen that code mixing is pervasive while people communicate over social media irrelevant of the mode being used. The fusion of languages makes it more challenging and requires consistent updates according to recent trends. The current paper addresses three approaches namely CRFs (Conditional Random Fields), Bi-LSTM (Long Short-term Memory) and CNNs( Convolutional Neural Networks). Firstly, for word-level language identification of code-mixed English-Punjabi text CRF based system uses lexical, contextual, character ngram, and special character features. Secondly, Recursive Neural Network namely Bi-LSTM with glove embedding is used for language identification and thirdly CNN with glove embedding is used for language identification. It is observed that CRFs is the best performing system with an f1-score of 0.96.

Keywords

Code Mixing, Language Identification, Deep Learning, Glove Embedding, Conditional Random Fields.

Full Text