People/Web Search Calendar Emergency Info A-Z Index UVA Email University of Virginia

Skip to Content

CS News Fall 2011

CS News:
Assistant Professor Jason Lawrence - Getting Underneath the Surface of Reality

Assistant Professor Jason Lawrence

The immense complexity of the natural world is daunting for researchers trying to create realistic computer-generated imagery. Minute surface imperfections, even a slight coating of dust, change the way light reflects off an object. Without these visual cues, a synthetic image can look like a BBC costume drama: too perfect to be real.

There are ways to approximate this complexity with existing three-dimensional modeling software, but given the amount of manual fine-tuning required, designers might well be tempted to exchange their mouse for a paintbrush. Assistant Professor Jason Lawrence has begun to break this impasse. He is developing novel techniques to acquire and process data that corresponds to an object’s material appearance as well as its geometry. “In a way, my work is about making it easier to create renderings that are more true to life because they are dustier and grittier,” he says.

Combining three-dimensional geometry and surface appearance is critical, for instance, for museums that are scanning their collections. “You can gain a richer, more flexible description that is impossible to achieve with a photograph,” Lawrence notes. “You can also use the data you acquire to digitally restore a damaged object or simulate the way it looked in different environments.”

Working with former graduate student Michael Holroyd (CS ’11), Lawrence has developed a coaxial optical scanner that captures shape and reflectance with a high degree of accuracy. He has also done work on processing the enormous datasets produced by image capture programs. Lawrence, who won a Faculty Early Development Program (CAREER) award from the National Science Foundation (NSF), has researched compressing these data sets so that they are more manageable, determining which light paths to sample, and developing intuitive ways for people to edit or manipulate the resulting digital models.

Recently, Lawrence was a co-principal investigator on a team that secured a $3 million grant from NSF — the largest in computer graphics in more than a decade — to build on this research and to apply his insights to threedimensional printing.