Broad Area of Research
- My primary area of research is image segmentation and their evaluation methods. A number of image segmentation schemes have been proposed by the researchers of computer vision community. However, I feel that sufficient effort has not yet not been put to evaluate those 100's and 1000's of algorithms. I also feel that image segmentation is a task that may not be generalized. Thus, it is unlikely that one algorithm can be used for all classes of images. One algorithm that produces satisfactory results for one set of images may not be as good as to other images classes.
- I am also focusing in object tracking algorithms, especially feature point based tracking schemes. My focus is on tracking objects without estimating motion at every point on the object. Such algorithms can reduce the processing time when on line computation is necessary.