Audio recording is a typical part of covert surveillance. However, the standard technologies used such as fixed microphone arrays, shotgun microphones and parabolic microphones are useful for picking up speech from distant speakers, but limited in their use by size and position constraints. A better understanding of microphone arrays with complex geometries could enable agents to place microphones at arbitrary positions in an environment such as a restaurant under tables, on lights, chairs, or on the clothing of agents in the room just minutes before the person under surveillance enters the room.
Vis Center faculty member, Kevin Donohue, has developed a project sponsored by the Federal Bureau of Investigation focusing on developing technology that combines wireless microphones mounted statically around an area or on moving platforms with clusters of computers to provide near-real time processing for tracking the suspect’s speech and delivering intelligible speech. The goal is to provide criminal justice and law enforcement agencies with the enhanced ability to covertly record and listen to remote conversation of suspect individuals in areas where multiple conversations are ongoing.
Researchers at the Vis Center recently published a paper demonstrating the system they have developed for LUT-based processing for real-time phase generation and 3-D reconstruction by means of structured light illumination. University of Kentucky student Kai Liu, along with fellow students, Yongchang Wang and Qi Hao as well as faculty members Daniel L. Lau and Laurence G. Hassebrook announced they have been able to use structured light illumination to produce real-time acquisition and creation of 3-D models.
Structured light illumination (SLI) is the process of projecting a series of light striped patterns onto an object. A digital camera then records the deformation in the pattern to reconstruct a 3-D model of the object’s surface. Using only a single camera and illumination source with a single processing computer, the system can easily be constructed from readily available parts.
In the past, SLI has not typically been used with video applications because of the very high frame rate needed. Based on the complexity of the calculations, video-based SLI systems have been required to record camera frames to memory and then apply off-line processing in order to reconstruct 3-D video.
This project has devised a lookup-table based solution that can generate real-time acquisition and display of 3-D scans. The system can acquire and display 3-D video in real-time. The potential uses for this technology include facial recognition applications, biometrics, and hand gesture recognition.
Today the EDUCE team is packing equipment for the trip home. Jurgen Stausens from SkyScan is retrieving the 1173 scanning system, securing it inside a very clever shipping container that will protect it for the truck ride back to Kontich, Belgium. The weight and sensitivity of the 1173 makes that process important, and SkyScan has developed a very reliable system for moving their equipment. The Kentucky team is organizing its shipment of computers, storage systems, and network infrastructure for the freight delivery company to send back to Lexington.
The project now moves into its final phase. The technical team will dig into the analysis of what has been collected. The media group will begin the final production schedule for the documentary film, which will tell the story of the EDUCE project and the broader, emerging world of “virtual conservation” and digital exploration.
Prior to the start of scanning, the EDUCE team consulted with Professor Daniel Delattre from the Sorbonne to learn about the typical structure of papyrus scrolls from this period. Normally the title page was at the end of the scroll, inside the tightest wraps. Sometimes the scroll was wound around a wooden core; other times there was an extra piece of papyrus that functioned as a support onto which the rest of the story was rolled. Two sections of papyrus were joined together at a seam called “kollesis”, allowing for a roll to be as long as the story required.
Knowing the specifics of the scrolls’ construction can help tremendously in the analysis – guiding our search for seams, finding the section which may likely contain writing to identify the title and the author, and establishing how the scroll was rolled up. For those interested in papyrology, check out the University of Michigan’s papyrus collection and their on-line description of some of the basics: http://www.lib.umich.edu/papyrus-collection/kollesis. Mapping these features to what can be seen in Micro-CT of papyrus is largely uncharted territory and gives the research team plenty of work to do.
The 79 AD eruption of Vesuvius doomed the Herculaneum scrolls to carbonization. Their discovery and excavation beginning in 1750 opened up a very fragile window into the only surviving library from classical antiquity. The four weeks of micro-CT scanning that the EDUCE research team has spent with just two of the 1,785 carbonized scrolls represents one month of the 23,160 months the scrolls have survived since their carbonization. If that total time were represented by the distance between the Institut de France in Paris, and Lexington KY (4,191 miles along a great circle on the earth), the month we have spent working with the scrolls would be like walking from the Institut to the Pyramid at the Louvre: about 1,000 feet, or a 5 minute walk.
However, in this short time we have peered past the damage inflicted by carbonization and the passage of time with the help of technology and digitization. As the research continues we will continue to learn how these new technologies can enable us to preserve these important artifacts of the past.
Requests for data can be made by contacting Brent Seales.