Home

Resume

Projects

Publications

Academics

Personal

Face recognition is a part of a wide area of pattern recognition technology. Recognition and especially face recognition covers a range of activities from many walks of life. Face recognition is something that humans are particularly good at and science and technology have brought many similar tasks to us. .Our project aims at identifying the faces of known individuals and can be used for authentication. We aim to combines both information theory and neural networks, integrating both eigenfaces, to find a compact representation, and neural networks, to classify the face images for a more practical face recognition system. The recognition is done by comparing characteristics of the new face to those of known individuals. First, face images are projected onto feature spaces that span the significant variations among face images. The significant features are known as (eigenfaces), because they are the eigenvectors (principal components) of the set of the faces. We then obtain the feature vectors of the face image that is presented as input for the neural network. The Neural Network gives significant advantages over the simple minimum distance classifiers used in eigenfaces recognition approach. The result of the recognition is deducted from the neurons of the output layer. Thus the identification is performed by using Multilayer Preceptor. Some advantages of this system are its simplicity, its rapidity and its ability for learning and generalization. To improve further we are aiming to replace multilayer preceptor by self training artificial neural networks.

                                                                                                   Live project for Kalptaru Investments, Indore

This project designs and implements database management software for a stock market brokerage firm, Kalptaru Investments. It includes development of database for maintaining the data for the share trading agency. The objectives of this project were to automate the manual system of maintaining registers, to provide a user-friendly interface, to provide effective storage of databases and to facilitate easy querying, retrieving and updating of records. It provides a user friendly interface for various transactions. Online help feature is also available. In all it is an efficient and easy way of share trading without actually in share market by using this software which can  be installed by trading agency. The whole project has been developed using ORACLE 8.0   as a backend tool and VISUAL BASIC 6.0 as a front end tool.

Character Recognition is a subset of Computer Vision where computer is able to interpret human handwriting and perceive it as an alphanumeric character instead of as an image. We have used template based technique for character recognition. It occurs in two phases. Firstly, in the teaching phase, characters are ‘taught’ to the system and each character is stored as a template in the computer’s database. It is implemented by user drawing the character in the drawing area using a mouse. The learning algorithm computes the bounding box across the drawn character and generates a bitmap font style template to store. It is possible to teach different versions of the same character. During the recognition phase the algorithm compares the bitmap template of test character to the internal database and gives the maximum match as output.

Steganography is the art and science of communicating in a way which hides the existence of the communication. The goal of Steganography is to hide messages inside other harmless messages in a way that does not allow any enemy to even detect that there is a second secret message present. This project has been developed software in Visual Basic 6 using the Least Significant Bit Insertion method. For each bit in the message, we pick a random pixel in the image and again randomly pick the pixel's red, green, or blue component. We then set the least significant bit in that component to 0 or 1 to encode the message bit Data we wish to embed consists of bytes (text, zips, encrypted data, executables: they're all byte-files).  This program simple spreads the bits of the data file among the least significant bits (or LSBs) of the color levels. This way, the color levels are only slightly altered.

This is an interactive tutorial developed in C for learning the various search techniques (DFS, BFS, A*, Hill Climbing etc). It has a pre- programmed graph for the introduction to search techniques. The user in the advanced stage can interactively feed an arbitrary graph for a particular search to be performed. A text display of the actions being performed at each node is also available. At the same time a queue, link-list, stack state is shown for understanding the memory management and implementation on a computer.

   AI Problem Implementation

Paint Express is a drawing tool that can be used to create simple or elaborate drawings. it provides graphic user interface enabling mouse pointer to select various menus and parameters . It enables saving and retrieving of files and also provides online help.

  Telephone Billing System

It was developed in COBOL for integrating all the aspects of telephone billing system right for managing customer’s database to keeping account of the telephone usage.

     Clinic Manager

Editor and Assembler