Projects

Satellite Remote Sensing Analysis Platform

Worked with Prof. Leland Pierce of the EECS department of University of Michigan on developing an open-source software platform for analysis of remotely sensed data, consisting of C++, Python and web libraries.
This library will include data storage and processing capabilities for satellite imagery, elevation maps, and include support for 2D and 3D models.
This project aims to develop a software that provides interoperability between all the different remote sensing systems and standards- something that has not been tried before in practice.

Developed a system for human identification based on iris matching.
Pictures of subjects were taken using a standard webcam to create a database. The iris was segmented out from these images using Canny edge detection and Hough transforms, followed by normalization using Daugman's rubber sheet model and feature extraction using Gabor filters.
These extracted features were then stored in the database as the templates to be tested against. Matching with test subjects were done on the basis of Hamming distance criterion.

Created a system capable of detecting various peaks(Q,S,T) of the ECG signal as well as detection of arrhythmia using wavelet transformation method in MATLAB.
This was done with the implementation of various adaptive and non-adaptive filter designs and noise removal techniques.
The detection of arrhythmia was performed using a series of threshold detection methods for the various peaks.

Developed a surveillance and monitoring system using RFID(Radio Frequency ID) tags implanted in VIT University ID Cards as a part of the SEDS(Students for the Exploration and Development of Space) Project Team.
These tags were used to determine the location of faculty and students carrying identity cards within the campus and relay them to a database server, using RFID readers.