Saurav Basu

Master's Graduate Student
School of Computing
University of Utah
(Aug 2004- Sep 2006)













   Research Projects:

  1. Noise Removal In Diffusion Tensor Images
    This was my Master's thesis research. We investigated techniques for filtering diffusion tensor images. Diffusion Tensor images are widely used to to study the tissue structure inside our brains. DT-MRI is one of 10 emerging technologies as of Mar/Apr 2006.Read about it here
    My advisor : Dr. P. T Fletcher
  2. Surface Reconstruction Project
    This project involved building a software tool to facilitate acquisition of 3D models of real world objects and segmenting them into surfaces which correspond to geometric primitives. These segmented surfaces can then be used to fit CAD models for use in reverse Engineering.

    Other projects I enjoyed working on :
  1. External Memory Sorter: (Algorithms Cs6962)
    Wrote a multi threaded external memory sorting program. Available physical memory 128 MB, file sizes to be sorted upto 16GB.

  2. Content Based Image Retrieval: (Computer Vision Cs6320)
    Prototyped a query by example image retrieval system based on Color Correlograms and Color Histograms and did a comaparative study on an image database of 1000 images from 10 different categories.Also investigated Texture Searches using Gabor filters.

  3. Image Mosaicing: (Image Processing Cs7966)
    Image mosaicing is stitching together images taken from different view points by aligning them together. i can use my program to build both planar and panaromic mosaics.

  4. Contrast Correction in Images:(Image Processing Cs7966) Studied and implemented contrast correction algorithms.Among them CLAHE (Contrast Limited Adaptive Histogram Equalization) gives the maximum flexibility in selectively increasing contrast in specific image regions.(No more low contrast photos!!)

  5. Feature Detection (Lines and Circles):(Image Processing Cs7966) Studied and implemented the Hough transform algorithm for line and circle detection.

  6. Dept From Stereo:(Undergrad Term Project) In this project i applied a stereo correlation technique to extract depth and visualize the result as a perspective view from any given view point.The idea is to mimic the human depth perception using a pair of cameras in place of the eyes.