Apurva A. Desai

Gujarati handwritten numeral optical character reorganization through neural network

This paper deals with an optical character recognition (OCR) system for handwritten Gujarati numbers. One may find so much of work for Indian languages like Hindi, Kannada, Tamil, Bangala, Malayalam, Gurumukhi etc, but Gujarati is a language for which hardly any work is traceable especially for handwritten characters. Here in this work a neural network is proposed for Gujarati handwritten digits identification. A multi layered feed forward neural network is suggested for classification of digits. The features of Gujarati digits are abstracted by four different profiles of digits. Thinning and skew-correction are also done for preprocessing of handwritten numerals before their classification. This work has achieved approximately 82% of success rate for Gujarati handwritten digit identification.

Segmentation of text lines into words for Gujarati handwritten text

A presentation on attempt to extract words from handwritten text lines in Gujarati script is hereby submitted. The very cursive nature of most Indian scripts makes the word extraction process a very critical one for Optical Character Recognition (OCR) activity. This cursive nature also causes difficulty during character extraction and modifier extraction. Word extraction is considered as one of the important stage of OCR, which directly affects the accuracy level of OCR. A combination of some proven methods like projection profile with morphological operations is used to enhance accuracy of the word extraction.

Zone Identification for Gujarati Handwritten Word

The research activity related to Optical Character Recognition (OCR) for almost all Indian languages needs more attention. The Gujarati language is no exception. This paper describes an important phase of OCR namely zone identification for Gujarati words. The zone identification is used to extract modifiers in the upper side of the base character called upper zone containing upper modifiers and lower side of the base character called lower zone containing lower modifiers. Detection of upper zone and lower zone will lead to detection of middle zone which contains most of the basic characters, conjunct characters and few modifiers, if any, as part of middle zone. The paper describes a process based on distance transform for identifying various zones for handwritten Gujarati words.

Extraction of Characters and Modifiers from Handwritten Gujarati Words

The research activity related to Optical Character Recognition (OCR) for almost all Indian languages is very less. Gujarati script is one of the scripts for which very less literature is available, as far as OCR activities are concerned. This paper describes one of the important phase of OCR, segmentation of handwritten words into its basic components namely basic characters, conjunct characters and modifiers, which are essential for recognition of a word. The paper describes methods for identification of zone boundaries for a word and usage of zone boundaries details for segmenting the word into its subcomponents. Connected component labeling is applied to detect subcomponents of a word, which can be further dissected if needed to obtain other subcomponents of word. It is the first attempt to dissect handwritten Gujarati words into its subcomponents.

Multi-layer Classification Approach for Online Handwritten Gujarati Character Recognition

In this paper, the authors present a multi-layer classification approach for online handwritten character recognition for the Gujarati characters. The Gujarati language consists of many confusing characters which lead to misclassification. Multi-layer classification technique is proposed to increase the accuracy of confusing characters. In the first layer of classification, SVM classifier with the polynomial kernel is used with all training data. If first layer classifier returns a character which can be confused with some characters than in the second layer, SVM with the linear kernel is used with confusing characters’ training data. A hybrid feature set consisting zoning features and dominant point-based normalized chain code feature is used in both layers of classification. The system is trained using a data set of 2000 samples and tested by 200 different writers. The authors have achieved an average accuracy of 94.13% with an average processing time of 0.103 s per stroke.

Human Computer Interaction Through Hand Gestures for Home Automation Using Microsoft Kinect

Gesture recognition has been an attractive area of research since a long time. With the introduction of Microsoft Kinect, hand gesture and body gesture recognition has become handy for the researchers. Here an innovative application has been presented which controls all electrical home appliances through hand gestures. The algorithm presented here is an assistive application useful for physically challenged and senior citizens. In this paper we have used Microsoft Kinect for image capturing along with some important computer vision (CV) and digital image processing techniques (DIP) for hand gesture recognition. Arduino Uno microcontroller and relay circuits are used for controlling electrical devices. The algorithm presented gives an accuracy of 88 %.