Also serving the communities of De Luz, Rainbow, Camp Pendleton, Pala and Pauma

Post-earthquake imaging detailed at San Diego conference

Since the Rochester Institute of Technology is located in upstate New York, the university’s imaging research for emergency response was expected to be used for wildfires, floods, or terrorist attacks. After the January 12 earthquake in Haiti killed approximately 212,000 people, displaced approximately 1.2 million survivors, and damaged or destroyed more than 30 percent of buildings in Port-Au-Prince, the RIT researchers were asked to help.

During the August 1 to 5 annual meeting of SPIE, the international optical society, David Messenger of RIT’s Center for Imaging Science presented “High-resolution and lidar imaging support to the Haiti earthquake relief effort” at the August 2 Spectral Methodologies and Applications session. The work was also performed by RIT researchers Jan van Aardt, Donald McKeown, and William Basener.

RIT received a National Science Foundation grant approximately a year prior to the SPIE annual meeting at the San Diego Convention Center. The purpose of the Information Products Laboratory for Emergency Response (IPLER) program was to provide information to first responders.

The system includes visible, short-wave infrared, medium-wave infrared, and long-wave infrared cameras along with a lidar (laser radar) system. The systems are integrated into a camera platform to allow for concurrent data collection.

Following the earthquake, the World Bank turned to RIT for damage assessment, beginning the process of sending the RIT team to Haiti. “We’re not set up for these sorts of things,” Messenger said.

By January 15 RIT had a verbal commitment, although not an official contract, from the World Bank for funding of field operations in Haiti. Because the electronics are military-grade, an export control license was needed to take the system out of the United States, and U.S. State Department officials worked through the holiday weekend (Martin Luther King Day was on January 18) to make the field work possible. Coordination of

Haitian airspace was also a logistical issue, although one easily resolved, and arrangements also had to be made for uploading the collected data.

The research team arrived in the Bahamas on January 19 and started flying over Haiti on January 21. In addition to the damage assessment work for World Bank, the U.S. Geological Survey asked the team to take high-resolution lidar images along the fault line, so that work was added to the data collection.

The data was taken from an altitude of approximately 2,500 feet. The visible imagery had a resolution of 15 centimeters while the infrared resolution was 83 centimeters. The ratio of five hits per square mile over the fault line provided for higher-resolution imagery of those features.

The aircraft refueling base was in the Dominican Republic while the research team’s actual base during the mission was in Aguadilla, Puerto Rico. The University of Puerto Rico is in Mayaguez, and arrangements were made to upload the data from that university. After the plane landed in the evening, the hard drives were pulled from the airborne instrumentation, taken to Mayaguez, and inserted into a university computer.

Flights took place for seven days, from January 21 through January 27, and totaled 1,933 line-miles (3,115 line-kilometers). A total of 148 flight lines covered approximately 250 square miles (630 square kilometers). The four cameras combined to take a total of 60,764 images, not including the lidar imagery.

Approximately 699 gigabytes of data were transmitted from the University of Puerto Rico. Although a standard Internet connection would have allowed for transmission of between one and two megabytes per second, advanced Internet2 capabilities allowed for transmission at 47 megabytes per second.

“We learned a lot about what we can and can’t do,” Messenger said. “We also learned about what people want in the field.”

In addition to assessing building damage, the mission also was able to locate internally displaced persons through the automated detection of blue tarps. “It was a very high-contrast target,” Messenger said.

While some displayed Haitians were forced into red tents, rusted roofs created more uncertainty about whether real-color red items were tents or roofs.

The RIT data is public, and as of Messenger’s presentation more than 20 terabytes of data have been downloaded from the RIT source. In addition to building damage assessment, fault line characteristics, and survivor locations, potential applications

include drainage patterns and rubble and debris assessment. “We’re very interested to see what other people can pull out of this data,” Messenger said.

To comment on this story online, visit


Reader Comments(0)

Rendered 06/16/2024 18:12