Prof. Tull will talk informally regarding his research areas of interest and discuss new directions.
Here's a short summary taken from his web page which describes some of his research goals:
In the near future, digital television, enhanced cable and wireless services will give consumers and professionals unprecedented access to video data in digital format. These emerging technologies will create a demand for new, more sophisticated tools for the manipulation and enhancement of digital image sequences, across a broad range of imaging modalities.
My research focuses on the development of these tools -- algorithms and systems that address the inevitable problems that impede the efficient acquisition, representation, and transmission of digital image sequences. These problems are often difficult, ill-posed problems that lack a unique and stable solution. An aspect of my research considers special (robust) procedures required to transform an ill-posed problem into a well-posed one. I have considered the used of these techniques in a broad range of image sequence problems including:
I am also interested in how constraints or cues from the human visual system impact the solution to these problems. My latest work shows that a special class of dynamic visual cues can be used successfully, in conjunction with robust statistical regularization, to obtain unique, meaningful solutions to the image restoration and motion estimation problems.
I am presently exploring new regularization techniques for ill-posed inverse problems, the efficient implementation of regularized iterative algorithms and intelligent/adaptive image acquisition systems.