Virtual Reconstruction of Classical Ruins via Image Processing
Departments: Electrical and Computer Science & Classical Civilizations
Advisor: Ismail Jouny
Classical Antiquity, that is in short Ancient Greek and Roman civilization, has left behind many structures as evidence of its rich history. Many of these structures are in ruins and have been unearthed and made available by archeologists, but few of these wonders are entirely intact. Often experts are able to determine what originally at a site but visualizing that site in its original state remains difficult for a layman. In light of this, this project strives to recreate the undamaged structures at a known site using 2D images combined with already established knowledge about that site and similar ruins. This is accomplished by treating each image with processing techniques and then using edge detection to identify broken elements and replace them with an appropriate restored element. Identification of damaged elements is accomplished by comparing a discovered object with a library of known sample objects extracted from similar pictures that were treated with the same image processing technique. Techniques explored were various kinds of filtering, segmentation, Hough and watershed transforms; images analyzed were primarily of a simplified model created for the project with most major elements present in a typical early Roman temple.
About Rob Lowery:
Rob Lowery is a senior at Lafayette College pursuing a B.S. in Electrical &Computer Engineering and an A.B. in Classical Civilization. He has done research bringing his two academic interests together under Dr. Ismail Jouny of the Lafayette ECE department during the 2015-16 academic year. Additionally, Rob is a member of the Varsity Swim Team and a leader in the Lafayette Disciplemaker’s Christian Fellowship. After graduation he will be working for NBC Universal Sports.