My main interest lies in the area of statistical image understanding with a focus on fundamental issues. This research area has also been called computational vision since one seeks to design computerized systems for understanding scenes from camera images, much like our own human vision system. There is a strong need for such tools in medical diagnostics, face recognition, video surveillance, undersea imaging, terrain mapping, and satellite image analysis. In military domain, the problem of battlefield target recognition continues to motivate newer research. In the area of homeland security, the need for this research is immense. There is a need for techniques for uniquely characterizing people using facial images, fingerprints, retinal scans, gait analysis, or a combination of the above. Another important area is medical image analysis using non-invasive imaging techniques such as MRI, ultrasound, and PET. With sophisticated imaging techniques, there is a great need for algorithms that analyze anatomical objects from observed images.

We treat the problem of image understanding as that of Bayesian statistical inference and address some fundamental issues: How can we mathematically represent objects of interest, say faces, in images? What approaches can be used to compare patterns of pixels in a statistical framework? Can we derive probability laws that govern the variability observed in images of interest? How can we statistically characterize recognition performance for a given set of images? What methods can be invented to quantify, compare, and analyze shapes of objects, such as human beings, in images?

Our research follows closely Grenander's school of pattern theory that advocates the following three steps:
1. Create representations in terms of algebraic systems with probabilistic superstructures
2. Analyze structures from the perspective of statistical inferences.
3. Develop efficient algorithms to facilitate applications.

For some of the recent ongoing projects, please refer to the page:
Statistical Shape Analysis and Modeling Group (SSAMG)
Home | Research | Resume | Teaching | Publications | Students | Service| News