You are on page 1of 8

SYNOPSIS

On the topic

Noisy Alphabet Recognization using Matlab with NN toolbox


Submitted to Electronics & Communication Engineering Department for The partial fulfillment of the degree

Bachelor of Technology
In

Electronics & Communication Engineering


Session: 2011-2012
Submitted by: Monisha Bhardwaj [0810931034] Neha Gupta [0810931408] Sonam Gautam[0810931056] Supervised by: Mr Brijesh kumar Sharma Asst.Prof., ECE Deptt. ACET , Aligarh

Aligarh College of Engineering & Technology Aligarh

List of Contents

1. Abstract 2. Introduction of matlab with NN tool box 3. Algorithm 4. Flowchart 5. Application 6. References

INTRODUCTION of Matlab WITH NN TOOLBOX

Neural Network Toolbox provides tools for designing, implementing, visualizing, and simulating neural networks. Neural networks are used for applications where formal analysis would be difficult or impossible, such as pattern recognition and nonlinear system identification and control. The toolbox supports feedforward networks, radial basis networks, dynamic networks, selforganizing maps, and other proven network paradigms. Key Features : Neural network design, training, and simulation Pattern recognition, clustering, and data-fitting tools Supervised networks including feedforward, radial basis, LVQ, time delay, nonlinear autoregressive (NARX), and layer-recurrent Unsupervised networks including self-organizing maps and competitive layers Preprocessing and postprocessing for improving the efficiency of network training and assessing network performance Modular network representation for managing and visualizing networks of arbitrary size Routines for improving generalization to prevent overfitting Simulink blocks for building and evaluating neural networks, and advanced blocks for control systems applications.

Matlab's Neural Network Toolbox (NNT) is powerful, yet at times completely incomprehensible. This is mainly due to the complexity of the network object. Even though high-level network creation functions, like newp and newff, are included in the Toolbox, there will probably come a time when it will be necessary to directly edit the network object properties. Part of my job is teaching a neural networks practicum. Since we wanted the students to concern themselves with the ideas behind neural networks rather than with implementation and programming issues, some software was written to hide the details of the network object behind a Matlab GUI. In the course of writing this software, I learned a lot about the NNT. Some of it I learned form the manual, but most of it I learned through trial-and-error. And that's the reason I wrote this introduction. In order to save you a lot of time I already had to spent learning about the NNT. This document is far from extensive, and is naturally restricted to my own field of application. Therefore, only feed-forward networks will be treated. I do think however that reading this can give you a firm enough background to start building your own custom networks, relying on the Matlab documentation for specific details. All Matlab commands given in this document assume the existence of a NNT network object named `net'. To construct such an object from scratch, type >> net = network; which gives you a `blank' network, i.e. without any properties. Which properties to set and how to set them is the subject of this document.

Algorithm

The algorithm for alphabet recognization is given as follows: [1] GUI Creation. [2] Define a matrix ALPHABET which contains the bit maps of the 26 letters of the alphabet. [3] Define target vectors TARGETS for each letter. Each target has 26 elements with all zeroes, except for a single 1. A has a 1 in the first element, B in the second. [4] Defining The Neutral Network. [5] Training the network Without Noise, setting training parameter [6] Start training :like [7] P=alphabet;T=targets;[net,tr]=train(net,P,T); [8] Training The Network with Noise. [9] Training The Second Network Without Noise. [10] Testing The Network. [11] Calculating Average Errors for 100 Sets of 26 Target Vectors. [12] Display Result

Applications
[1] CoEvolution of Neural Network for control of Pursuit & Evasion. [2] Learning the Distribution of Object Trajectories for Event Recognition. [3] Radiosity for Virtual Reality Systems. [4] Autonomous Walker & Swimming Eel. [5] Robocup: Robot World Cup. [6] Using HMMs for Audio-to-Visual Conversion. [7] Artificial Life:Galapagos. [8] Speechreading(Lipreading) [9] Detection and Tracking of Moving Targets. [10] Real-time Target Identification for Security Application. [11] Facial Animation. [12] Behavioral Animation and Evolution of Behavior. [13] A Three Layer Feedforward Neural Network. [14] Artificial Life for Graphics, Animation, Multimedia, and Virtual Reality: Siggraph 95 Showcase [15] Creatures:The World Most Advanced Artificial Life!

References

[1] Bhoopendra Kumar Sharma, Devendra Singh Bharanger and Brijesh Kumar Sharma CHARACTER RECOGNITION USING ANN national conferences on emerging technologies in VLSI,135-153,18 September-2011. [2] I.M.Qureshi and A.Jalil, OBJECT RECOGNITION USING ANN" with backpropogation Alogrithm , Pakistan Journal of the applied. [3] Srinivasa Kumar Devireddy,Settipalli Appa Rao, HAND WRITTEN CHARACTER RECOGNITION using back propagation Network, Journal of Theoretical and Applied Information Techology,

2005-2009JATIT,pp.256-269,www.jatit.org. [4] Shashank Aroakar, "VISUAL CHARACTER RECOGNITION using Artificial Neutral Network, University Of Mumbai, india,pp.1-7. [5] P.K.Dash, P.K.Nanda,Saha and R.Doariswami, ARTFICIAL

NEURAL NETWORK & PATTERN RECOGNITION Approach for narrowband signal Extraction, 91TH0374-9/91/0000

0288501.00@1991 IEEE,pp.288-292.

ABSTRACT

The research aim at designing and developing a Neural Network based Character recognition system and capable of recognition a character correctly from noise.Here it is assumed that received character X is with a user defined noise level. When the user presses Start Recovery, the training of neural network starts and finally the input with noisy version of X, and correct version of X is received.

You might also like