NEWS

3-D computer program helps athletes improve techniques, can even predict performance

By JOY BUCHANAN
UNC School of Medicine

CHAPEL HILL -- Though a well-timed discus throw or the powerful leap of a successful long jump might seem the product of muscle strength and pure athletic talent, a University of North Carolina at Chapel Hill researcher says each actually is the product of mechanics, geometry and science.

Dr. Bing Yu has developed a software program that captures an athlete’s every move and turns it into three-dimensional computer animations that can help the athlete throw farther or jump longer distances.

"We analyze the biomechanics of their techniques to see what the problems are and how they can improve their performance," said Yu, associate professor of physical therapy at UNC’s School of Medicine and a member of the university’s Center for Human Movement Science.

Yu developed the computer program for the elite discus throwers of USA Track & Field, the national governing body for track and field, long-distance running and race walking. He uses two high-speed cameras to videotape an athlete during training or a competition. The videos are calibrated and digitized. Once he combines the data from the two cameras, that information is put into a computer program that generates a 3-D image of the athlete’s movements.

Yu’s animation program will soon have a prominent spot in the new National Track & Field Hall of Fame in New York City. The animation was selected for the exhibit, "What Makes A Champion?" Visitors to the museum will watch 3-D figures glide across a screen, displaying the biomechanics of world-class athletes.

"It’s a beautiful piece," said Keats Myer, project manager for the hall of fame. The animation will help visitors, especially children, understand how their movements influence their performance, she said. The grand opening for the hall of fame and biomechanics exhibit is scheduled for January 2004.

Yu’s 3-D movement images are black-and-white polygonal figures that jump, turn and throw as the athletes do. The figures allow Yu and the athletes to see the exact placement of critical body landmarks, such as feet, shoulders and hips. The image can be tilted or rotated to view from infinite angles.

Yu can use the image to find the technique or motion factor most strongly related to an athlete’s performance, he said. Yu also can compare the movements of two different athletes to determine what an athlete can do to mimic the performance of a better athlete.

The computer animation allows athletes and trainers to use science where they once relied on trial and error and a lot of luck, he said.

"Athletes used to guess how to improve their techniques. You could say if you change this, maybe you’ll have a better outcome. But now we can pinpoint where the problem is. Our program makes the whole process more scientific."

And it works.

In November 2001, Olympic discus thrower Suzy Powell’s best throw was 65 meters. She went through the biomechanical analysis process with Yu, and he made recommendations specific to her techniques. Just five months later, in April 2002, at the University of California at San Diego Open,

Powell threw a record 69.4 meters or 227 feet. That was 25 centimeters better than the record set in 1986.

"My first non-reverse throw was around 217 feet. My second throw was a reverse at 227," Powell said. "His evaluation of my potential distance was right."

Yu can also predict an athlete’s performance using his recommendations. At the 1992 Barcelona Olympics, Yu and his team predicted that the winning men’s triple jump would be about 18.30 meters. Mike Connolly, a U.S. athlete who used the researcher’s recommended technique, won the event with a jump of 18.25 meters. Yu also predicted the winning men’s triple jump at the 1996 Atlanta Olympics would be 18.11 meters. Kenny Harrison, a U.S. athlete, won with an 18.09 jump.

Yu said understanding how we move is not just important to athletes. The animation program will help UNC researchers examine movement for medical treatment and therapy.

"Our understanding of human movement is very limited," he said. "We are studying human movement right now to determine how to use movement to prevent injuries and treat injuries."

- 30 -

Note: Contact Yu at (919) 843-8643 or byu@med.unc.edu.
School of Medicine contact: Les Lang, (919) 843-9687 or llang@med.unc.edu