Mayuresh Kulkarni (मयुरेश कुलकर्णी)

Department of Electrical Engineering, University of Cape Town.
Supervisor : Dr. Fred Nicolls

Email address : mayuresh.kulkarni@uct.ac.za, mayuresh87@gmail.com

Physical address :
Room 616, Menzies Building,
Digital Image Processing Group,
University of Cape Town,
Rondebosch, 7701.
South Africa.

Research Interests:

Research Interests
Digital image processing, medical imaging, 3D reconstruction, image segmentation, multivariate statistical modelling. Other interests include database management and web design.

PhD Research - Monte-Carlo simulation of slot-scanning X-ray imaging systems (2011-2014)
Simulation of slot-scanning X-ray imaging systems, full-body X-ray systems, mammography and tomosynthesis systems. The simulation tool uses the Monte-Carlo implementation PENELOPE to simulate the X-ray photon transportation problem. Modulation Transfer Function (MTF), Noise Power Spectrum (NPS) and Detective Quantum Efficiency (DQE) are used to validate the simulations. Tomosynthesis is simulation using angular intervals, angular range and number of projections and their effect on the resulting reconstructions is studied. Filtered backprojection is used to reconstruct the 2D X-ray projections. (PhD Thesis [pdf])

Masters Research - Image segmentation using Graph cuts (2009-2010)
My current research involves interactive image segmentation using graph cuts. The image is viewed as a graph with the pixels, or group of pixels, as nodes and the connections between them as edges. Min-cuts is used to segment each pixel into foreground or background. Various ways of setting up the edge weights is studied and the performance is evaluated using precision-recall curves and F-score. (Masters Thesis [pdf])

Undergraduate Thesis - X-ray Image Segmentation using ASMs (July-November 2008)
Investigated the performance of various image processing techniques on X-ray images. Studied Active Shape Models (ASMs) and applied them to segment bone from X-ray images. The algorithms implemented take user input in the form of training X-ray images and then tracks the shape in a test image. The performance of the algorithm, with varying the input parameters, was evaluated. The project involved writing a thesis paper and presenting it to a panel of experts.
(Undergraduate Thesis [pdf] and [presentation])


Publications:

M. Kulkarni and F. Nicolls. "Graph cuts with shape priors for segmentation". Twenty-first Annual Symposium of the Pattern Recognition Association of South Africa (PRASA2010), Stellenbosch,
South Africa, 2010. [paper]

M. Kulkarni and F. Nicolls. "Interactive Image Segmentation using Graph Cuts". Twentieth Annual Symposium of the Pattern Recognition Association of South Africa (PRASA2009), Stellenbosch,
South Africa, 2009. [paper] [presentation]


Last Modified -- Mayuresh Kulkarni (January 2015)