Entity

Time filter

Source Type

New Delhi, India

Choudhary A.,Maharaja Surajmal Institute | Rishi R.,TITS | Ahlawat S.,Maharaja Surajmal Institute of Technology | Dhaka V.S.,IMS
2010 The 2nd International Conference on Computer and Automation Engineering, ICCAE 2010 | Year: 2010

Aim of this paper is to analyze the performance of back-propagation feed-forward algorithm using various different activation functions for the neurons of hidden and output layers. For sample creation, 250 numerals were gathered form 35 people. After binarization, these numerals were clubbed together to form training patterns for the neural network. Network was trained to learn its behavior by adjusting the connection strengths at every iteration. The conjugate gradient descent of each presented training pattern was calculated to identify the minima on the error surface for each training pattern. Experiments were performed by selecting different combinations of two activation functions out of the three activation functions 'logsig', 'tansig' and 'purelin' for the neurons of the hidden and output layers and the results revealed that the percentage recognition accuracy of the neural network was observed to be optimum when 'tansig'-'tansig' combination of activation functions was used for neurons of hidden and output layers. ©2010 IEEE. Source


Choudhary A.,Maharaja Surajmal Institute | Rishi R.,Maharshi Dayanand University | Ahlawat S.,Maharaja Surajmal Institute of Technology
Procedia Computer Science | Year: 2013

Characters extraction is the most critical pre-processing step for any off-line text recognition system because the characters are the smallest unit of any language script. The paper proposes an approach to segment character images from the text containing images and computer printed or handwritten words. This segmentation approach is based on a set of properties for each connected component (object) in the whole binary image of the machine printed or handwritten text containing some other images. These words which are printed along with some images are of different lengths and are printed by different cursive fonts of different sizes. This character extraction technique is applied for the segmentation of untouched characters from the machine printed or handwritten words of varying length written on a noisy background having some images etc. Very promising results are achieved which reveals the robustness of the proposed character detection and extraction technique. © 2013 The Authors. Published by Elsevier B.V. Source


Choudhary A.,Maharaja Surajmal Institute | Ahlawat S.,Maharaja Surajmal Institute of Technology | Rishi R.,Maharshi Dayanand University
Studies in Fuzziness and Soft Computing | Year: 2015

The feature extraction is one of the most crucial steps for an Optical Character Recognition (OCR) System. The efficiency and accuracy of the OCR System, in recognizing the off-line printed characters, mainly depends on the selection of feature extraction technique and the classification algorithm employed. This chapter focuses on the recognition of handwritten characters of Roman Script by using features which are obtained by using binarization technique. The goal of binarization is to minimize the unwanted information present in the image while protecting the useful information. Various preprocessing techniques such as thinning, foreground and background noise removal, cropping and size normalization etc. are also employed to preprocess the character images before their classification. A multi-layered feed forward neural network is proposed for classification of handwritten character images. The difference between the desired and actual output is calculated for each cycle and the weights are adjusted during error back-propagation. This process continues till the network converges to the allowable or acceptable error. This method involves the back propagation-learning rule based on the principle of gradient descent along the error surface in the negative direction. Very promising results are achieved when binarization features and the multilayer feed forward neural network classifier is used to recognize the off-line cursive handwritten characters. © Springer International Publishing Switzerland 2015. Source


Choudhary A.,Maharaja Surajmal Institute | Rishi R.,Maharshi Dayanand University | Ahlawat S.,Maharaja Surajmal Institute of Technology
Advances in Intelligent and Soft Computing | Year: 2012

The recognition accuracy of an Optical Character Recognition (OCR) system mainly depends on the selection of feature extraction technique and the classification algorithm. This paper focuses on the recognition of handwritten digits using projection profile features. Vertical, Horizontal, Left Diagonal and Right Diagonal directions are the four different orientations that are used for abstracting features from each handwritten digit. A feed forward neural network is proposed for recognition of digits. 750 digit samples are collected from 15 writers; each writer contributed each of the 10 digits 5 times. Thus a local database containing 750 digit samples is used for training and testing of the proposed OCR system. Preprocessing of handwritten digits is also done before their classification. The combination of proposed feature extraction method along with back-propagation neural network classifier is found to be very effective as it yields excellent recognition accuracy. © 2012 Springer-Verlag GmbH Berlin Heidelberg. Source


Choudhary A.,Maharaja Surajmal Institute | Rishi R.,Maharshi Dayanand University
Advances in Intelligent Systems and Computing | Year: 2014

This paper is focused on evaluating the capability of MLP and RBF neural network classifier algorithms for performing handwritten character recognition task. Projection profile features for the character images are extracted and merged with the binarization features obtained after preprocessing every character image. The fused features thus obtained are used to train both the classifiers i.e. MLP and RBF Neural Networks. Simulation studies are examined extensively and the proposed fused features are found to deliver better recognition accuracy when used with RBF Network as a classifier. Source

Discover hidden collaborations