Rajib Ghosh

Study of two zone-based features for online Bengali and Devanagari character recognition

This paper presents two zone-based feature extraction approaches for online handwritten character recognition of two major Indic scripts-Bengali and Devanagari. Here, each stroke of an online character is divided into a number of local zones. In the first approach, named Zone wise structural and directional features (ZSD), structural and directional features are extracted for each stroke in each of these local zones. In the second approach, named Zone wise slopes of dominant points (ZSDP), the dominant points are detected first from each stroke and next the slope angles between consecutive dominant points are calculated and features are extracted in these local zones. Next, these features are fed to SVM classifier for stroke recognition. The constituent stroke combinations of characters are matched with training data and characters are recognized accordingly. Using ZSD, the recognition performances for Bengali (9,800 test data) and Devanagari (10,000 test data) scripts are 87.48% and 85.10% and with ZSDP, the accuracies are 92.48% and 90.63% respectively.

Devanagari text extraction from natural scene images

In scenic images, information in the form of text provides vital clues for most applications based on image processing. These include assisted navigation content based image retrieval, automatic geocoding and understanding the scene. But in a multicolored complex background, it is quite a daunting task to locate the text. This task is daunting because of non-uniformity in illumination, complexity of the backdrop, and differences in the size font & line-orientation of the text. We propose a novel approach for Devanagari text extraction from natural scene images in this paper. We can use a text-to-speech engine or Optical Character Reader to recognize the extracted text. The basis of our scheme is to analyze the CCs. This is done to extract Devanagari text from scenic images captured by camera. The presence of head line is unique to this script. Our scheme makes use of mathematical morphological operations to extract the headlines. Also the binarization of scenic images was studied. Here the effectiveness of the adaptive thresholding approach was observed. The algorithm was tested on Devanagari text contained within a collection of 100 scenic images.

A novel feature extraction approach for online Bengali and Devanagari character recognition

This paper presents an online handwritten character recognition system for two major Indic scripts-Bengali and Devanagari. In this proposal, a novel approach for feature extractions is described in which each online stroke information of a character is divided into a number of local zones. For each online stroke information different structural and directional features are extracted separately in each of these local zones. Next, these features are concatenated and fed to SVM classifier for recognition. The character recognition accuracy obtained is 87.48% for Bengali script and 84.10% for Devanagari script on 4900 and 5000 test samples respectively.

Segmentation of Online Bangla Handwritten Word

To take care of variability involved in the writing style of different individuals in this paper we propose a robust scheme to segment unconstrained handwritten Bangla words into characters. Online handwriting recognition refers to the problem of interpretation of handwriting input captured as a stream of pen positions using a digitizer or other pen position sensor. For online recognition of word the segmentation of word into basic strokes is needed. For word segmentation, at first, we divide the word image into two different zones. The upper zone is taken as the 1/3rd of the height of the total image. Now, based on the concept of downside movement of stroke in this upper zone we segment each word into a combination of basic strokes. We segment at a pixel where the slope of six consecutive pixels satisfies certain angular value. We tested our system on 5500 Bangla word data and obtained 81.13% accuracy on word data from the proposed system.

Comparison of Zone-Features for Online Bengali and Devanagari Word Recognition Using HMM

This paper presents a comparative study of three feature extraction approaches for online handwritten word recognition of two major Indic scripts-Bengali and Devanagari using Hidden Markov Model (HMM). First approach uses feature extraction from whole stroke without local zone division after segmenting the word into its basic strokes. Whereas, other two approaches consider the segmentation of a word into its basic strokes and a local zone wise analysis of each online stroke. Among these two zone wise local features, one takes into account structural and directional features and other uses dominant points, detected from strokes using slope angles, to find the local features. These features are studied in HMM-based word recognition platform. From the comparative study of the word recognition results, we have noted that dominant point based local feature extraction provides best accuracies for both Bengali and Devanagari scripts. We have obtained 90.23% and 93.82% accuracies for Bengali and Devanagari scripts respectively.

Stroke segmentation of online handwritten word using the busy zone concept

To take care of variability involved in the writing style of different individuals a novel approach has been proposed in this article to segment unconstrained handwritten Bangla words into characters. Online handwriting recognition refers to the problem of interpretation of handwriting input captured as a stream of pen positions using a digitizer or other pen position sensor. For online recognition of word the proper segmentation of word into basic strokes is very much important. For word segmentation, at first the busy zone of the whole word is calculated and then an estimated headline is imagined just above the starting point of the busy zone. Remove all the pixels crossing the estimated headline by checking their distance. Finally the segmentation is done. The system has been tested on 5500 Bangla word data and obtained around 94.9% of correct segmentation on word data from the proposed system.

A Novel Approach of Skew Correction for Online Handwritten Words

Segmentation of a word into basic characters or strokes is an essential and necessary step for character recognition in many handwritten word recognition systems. The one of the major difficulties in character segmentation is the existence of the skew in the handwritten word. But skew detection in the bangla handwritten word is difficult because of its shape variability of the characters as well as larger number of character classes. If the skew correction is done successfully then the character segmentation of the word will be more perfect and as a consequence the percentage of the correct word recognition will be higher. In this paper, we propose a novel method for skew detection and skew correction of online Bengali handwritten word through holistic approach. This approach works based on center of gravity of left part and right part of a handwritten word. After finding the center of gravity it calculates the angle of the line which connected the two gravity centers in relation to horizontal line. Then Rotates the word clockwise by the angle θ if θ 90o. All the pixel moves to the particular angle to correct the skew. The algorithm has been verified on a database of 3000 Bengali word data collected from different people of different age group and it gives 92.22% accuracy on word data from the proposed system. General Terms Pattern Recognition.

New Algorithm for Skewing Detection of Handwritten Bangla Words

Segmentation of a word into basic characters or strokes is an essential and necessary preprocessing step for character recognition in many handwritten word recognition systems, especially in case of handwritten bangla words. The major difficulty in character segmentation is the cursive script. This is because different person have different styles for their handwriting. Here, in this article a novel approach for skew detection followed by skew correction has been presented for online handwritten Bangla words. Here, we have used a slight variation of the projection profile method to calculate the amount of skew in an online Bangla handwritten word. The algorithm has been verified on a database of words collected from different people.

Study of Zone-Based Feature for Online Handwritten Signature Recognition and Verification in Devanagari Script

This paper presents one zone-based feature extraction approach for online handwritten signature recognition and verification of one of the major Indic scripts–Devanagari. To the best of our knowledge no work is available for signature recognition and verification in Indic scripts. Here, the entire online image is divided into a number of local zones. In this approach, named Zone wise Slopes of Dominant Points (ZSDP), the dominant points are detected first from each stroke and next the slope angles between consecutive dominant points are calculated and features are extracted in these local zones. Next, these features are supplied to two different classifiers; Hidden Markov Model (HMM) and Support Vector Machine (SVM) for recognition and verification of signatures. An exhaustive experiment in a large dataset is performed using this zone-based feature on original and forged signatures in Devanagari script and encouraging results are found.