Archaeology

scanning_animation_100x100.gif

 

 

 

 

 

A central theme of the laboratory work is interdisciplinary collaboration. One of the strongest interdisciplinary collaborations is with archaeologists and anthropologists from several institutions around the world. Work in this regard focuses on how newly developed digital technologies such as 3D scanning and digital imagery and computational techniques associated with these technologies can contribute to solving difficult problems in archaeology and anthropology.

The castle reconstruction project explores new ways for measuring, modeling, and estimating ancient structures using the ruins of a Crusader fortress, Apollonia-Arsuf, as a case example. Novel measurement devices have been developed that empower archaeologists to create 3D surface models by integrating a dense collection of (x,y,z) surface measurements from in-situ architectural remains with digital photographs. Another aspect of the project develops a specialized computer language, which allows users to specify shape grammar, enabling researchers to compactly represent complex architectural structures in terms of simpler sub-structures, e.g., the wall of a building consists of floors, each floor consists of brick and mortar with facade features inserted at specific locations such as windows. By changing the variables in the computerized shape grammar, different variations of an architectural structure can be automatically generated. The final part of the project seeks to automatically or semi-automatically reconstruct damaged and collapsed architecture from 3D scan data and a user-specified shape grammar. Here, the measured 3D scan data is used to constrain some or all of the variables of the shape grammar program in an effort to automatically or semi-automatically reconstruct the most likely structure of the damaged architecture. Work on this project is supported through NSF Grant NSF-IIS 0808718.

The following links provide access to many of the 3D models that have been used in our publications.

Fragments for Pottery Vessel Reconstruction from 3D Scans of the Vessel Fragments

The following archaeological fragments were used to develop methods capable of automatically reconstructing a broken pot from 3D scans of the outer surface of its fragments.

A Simple Pot Broken For the Purposes of Experimentation

This pot was broken in the laboratory by dropping the pot from a height of approximately 4 feet onto a tiled floor. The broken pot consisted of many fragments. However, due to chipping and a there are only 13 fragments of sufficient size for scanning the resulting reconstruction will exhibit a small hole at the point of impact where several small fragments were generated that were not scanned.

p1.obj

p2.obj

p3.obj

p4.obj

p5.obj

p6.obj

p7.obj

p8.obj

p9.obj

p10.obj

p11.obj

p12.obj

p13.obj

Related Articles:

  1. Willis, A. and Cooper, D., Alignment of Multiple Non-Overlapping Axially Symmetric 3D Datasets, International Conference on Pattern Recognition (ICPR), Vol. IV, pp. 96–9, 2004.
  2. Willis, A. and Cooper, D., Bayesian Assembly of 3D Axially Symmetric Shapes from Fragments, Conference on Computer Vision and Pattern Recognition (CVPR), Vol. I, pp. 82–89, 2004.
  3. Willis, A. and Cooper, D. B., and Andrews, S. and Baker, J. and Cao, Y. and Han. D. and Kang, K. and Kong, W. and Leymarie, F. and Orriols, X. and Velipasalar, S. and Vote, E. and Joukowsky, M. S. and Kimia, B. and Mumford, D., Bayesian Pot-Assembly from Fragments as Problems in Perceptual-Grouping and Geometric-Learning, International Conference on Pattern Recognition (ICPR), Vol. III, pp. 297–302, 2002.
  4. Cooper, D. B. and Willis, A. and Andrews, S. and Baker, J. and Cao, Y. and Han, D. and Kang, K. and Kong, W. and Leymarie, F. and Orriols, X. and Velipasalar, S. and Vote, E. and Joukowsky, M. S. and Kimia, B. and Laidlaw, D. and Mumford, D., Assembling Virtual Pots from 3D Measurements of their Fragments, VAST International Symposium on Virtual Reality Archaeology and Cultural Heritage, pp. 241–253, 2001.

 

A Real-World Archaeological Vessel: A 7-Fragment Nabatean Drinking Vessel

This vessel was discovered at the Brown University excavation of the Great Temple, a structure in the ancient city of Petra in Jordan. There are 7 fragments available for the vessel which form the sides of the drinking vessel, the bottom of the vessel remains undiscovered to date.

petra1867.obj

petra1869.obj

petra1875.obj

petra1899.obj

petra1901.obj

petra1913.obj

petra1938.obj

Related Articles:

  1. Willis, A. and Cooper, D. B., From Ruins to Relics: Computational Reconstruction of Ancient Artifacts, IEEE Signal Processing Magazine, Vol. 25, No. 4, pp. 65-83, July 2008.
  2. Sui, Y. and Willis, A., Using Markov Random Fields and Algebraic Geometry to Extract 3D Symmetry Properties, Fourth International Symposium on 3D Data Processing, Visualization and Transmission (3DPVT), June 18-20, Atlanta, GA, 2008.
  3. Willis, A. and Cooper D. B., Estimating a-Priori Unknown 3D Axially Symmetric Surfaces from Noisy Measurements of their Fragments, Third International Symposium on 3D Data Processing, Visualization and Transmission (3DPVT), pp. 334-341, 2006.

 

Fragments for Pot Vessel Axis Estimation from a 3D Scan of  a Vessel Fragment

The following archaeological fragments were used to develop methods capable of accurately estimating the central axis and profile curve of a pottery vessel given a scan of the outer surface of a fragment from the vessel.


p642.obj

p654.obj

p967.obj

p997.obj

p1135.obj

p1313.obj

Related articles:

  1. Willis, A., Stochastic 3D Geometric Models for Classification, Deformation, and Estimation, Ph.D. Thesis, Brown University, May 2004.
  2. Willis, A. and Cooper, D. B. et. al., Accurately Estimating Sherd 3D Surface Geometry with Application to Pot Reconstruction, Conference on Computer Vision and Pattern Recognition (CVPR) Workshop, 2003.
  3. Willis, A. and Orriols, X. and Velipasalar, S. and Cooper, D. B., Extracting Axially Symmetric 3D Geometry from Limited 3D Range Data, Technical Report LEMS-192, Brown University, Providence, RI, 2001.

 

Shared goals, explored through trial and error, have led disparate research and educational organizations to work together in productive modes, jointly where desirable, individually where possible. The Brown University, Division of Engineering, Laboratory for Man/Machine Systems (LEMS), the Brown University Joukowsky Institute for Archaeology and the Ancient World, the nonprofit educational outreach Institute for the Visualization of History, Williamstown, MA, and archaeologists at Tel Aviv University, Israel, in recent years have established a working relationship. The results have been important accomplishments in both computervision/pattern-recognition (CVPR) research in core-mainstream and new directions, as well as in new digital archaeology research directions. This proposal will result in: major contributions to core CVPR research; an interactive data-collection/visualization/geometric information-extraction/results-storage system; and major paradigmshifts in archaeological research concepts, methodologies, and results.


The grant focuses on four highly interdependent categories: 1) Semiautomatic development of an internet accessible archaeological space-time excavation-site database containing both traditional information about findings, and images, video, and extracted 3D representations and geometry; 2) Automatic and Semiautomatic 3D object, fragment, and large architectural chunk surface and geometry estimation from moving and multiple stationary image and video cameras; 3) Automatic and semi-automatic re-assembly of objects, especially glass or ceramic, and large architectural structures and sites, specifically the Crusader Castle at Apollonia, Israel, from estimated 3D representations of their many fragments or architectural chunks; 4) Interactive tools and 3D immersion for the purpose of interacting with the database, manipulating fragments, objects, and architectural substructures to facilitate object and architectural structure re-assembly, making precise metric measurements at an archaeological site from remote locations, and facilitating collaborative research. Included in these is CVPR search tools for artifacts, objects or architectural structures based on decorative painting on the surface or 2D or 3D geometry, both within the database and across databases at different locations. The space-time database, populated by images and video taken by moving and stationary cameras, as well as reconstructed 3D shape provides a detailed reviewable history of site excavation, and also the results of subsequent processing of this data, and walkthroughs the user wants to save. Our work will focus on, but not necessarily be limited to, the Apollonia site in Israel (on the UNESCO list of the hundred most endangered sites).


Intellectual Merit: The preceding constitute a complete integrated system of four highly interdependent categories. Categories 2 and 3 are completely research, and 1 and 4 are partially research. They are all intellectual explorations individually, but more importantly developing this system as an integrated whole results in a system of exceptional power, one that is significantly more powerful, conceptually and functionally, than the sum of its four parts. As a system, the categories are developed from more global and fundamental perspectives, with better tools, and with an important, exciting test environment.

Science and Broader Impacts: The benefits to the discipline of archaeology are multiple and persuasive. These developing tools allow unprecedented access to and analysis of past human activity, with geometry providing the capability to recognize, visualize, and evaluate forms of material culture accurately, rapidly and (above all) non-destructively. The results of this grant will benefit not only the present applicants, but archaeological studies worldwide. The vital areas of Information Science benefitting directly from this research are: geometric learning, 2D and 3D shape theory, geometric inferencing from multiview images and video, pattern recognition and database design. The algorithms and software applicable to other applications, e.g. homeland security, remote collaboration, education, and entertainment. The clear potential for developing educational tools for students already attracted to archaeology, is also very compelling, and to begin this process we highlight one of our three specific math/archaeology projects for high school students for the purpose of exposing them to and exercising them on application of certain mathematical topics to certain archaeology site tasks. The engineering team has made major contributions to these areas, is currently heavily involved in this research, and will be initiating new directions from our past experience. The members of the archaeology team are distinguished researchers-educators who have been involved in aspects of computational archaeology, are heavily involved in excavation at Apollonia and interpretation there and elsewhere, and are eager to push in the new directions that the group has jointly proposed.

Total: $2.6M

Sub: $305k

Using Markov Random Fields and Algebraic Geometry to Extract 3D Symmetry Properties

Presented at 3DPVT 2008 June 18-20, 2008 Atlanta, Georgia

Fourth International Symposium on

3D Data Processing, Visualization, and Transmission

Yunfeng Sui and Andrew Willis (UNC-Charlotte, USA)

 

 

In this paper, we present a new technique for solving the difficult problem of estimating the axis of symmetry for axially-symmetric surfaces. Accurate solutions to this problem are important in archaeology for systems that seek to reconstruct pottery vessels from measurements of their fragments. Our approach estimates quadratic surfaces at each measured surface point and uses a Markov Random Field superimposed on the measured surface mesh to estimate a collection of surface patches, each of which lies close to a single 3D quadratic surface. For each surface patch we estimate an quadratic implicit polynomial whose coefficients directly provide an estimate of the unknown axis location and orientation. Competing estimates of the global axis are combined using a Maximum Likelihood Estimation (MLE) framework that reflects the uncertainty present in the estimates computed from each surface patch. Our approach differs from past approaches by combining estimates derived from large surface regions that include many measurements instead of combining many local (often pointwise) estimates of the surface to determine the global estimate. Estimates from these large regions are more robust to noise and have sufficient data to generate statistics that accurately reflect the uncertainty in the computed estimates. As such, each estimate of the central axis is less susceptible to outliers and the overall axis estimate is significantly improved.

Using Markov Random Fields and Algebraic Geometry to Extract 3D Symmetry Properties

Presented at 3DPVT 2008 June 18-20, 2008 Atlanta, Georgia

Fourth International Symposium on

3D Data Processing, Visualization, and Transmission

Yunfeng Sui and Andrew Willis (UNC-Charlotte, USA)

In this paper, we present a new technique for solving the difficult problem of estimating the axis of symmetry for axially-symmetric surfaces. Accurate solutions to this problem are important in archaeology for systems that seek to reconstruct pottery vessels from measurements of their fragments. Our approach estimates quadratic surfaces at each measured surface point and uses a Markov Random Field superimposed on the measured surface mesh to estimate a collection of surface patches, each of which lies close to a single 3D quadratic surface. For each surface patch we estimate an quadratic implicit polynomial whose coefficients directly provide an estimate of the unknown axis location and orientation. Competing estimates of the global axis are combined using a Maximum Likelihood Estimation (MLE) framework that reflects the uncertainty present in the estimates computed from each surface patch. Our approach differs from past approaches by combining estimates derived from large surface regions that include many measurements instead of combining many local (often pointwise) estimates of the surface to determine the global estimate. Estimates from these large regions are more robust to noise and have sufficient data to generate statistics that accurately reflect the uncertainty in the computed estimates. As such, each estimate of the central axis is less susceptible to outliers and the overall axis estimate is significantly improved.


The paper was presented by
Yunfeng Sui

This video provides details on custom-designed software that takes as input 3D (x,y,z) measurements of the outer surfaces of a set of broken pottery fragments and outputs the "most likely" pots from which the fragments were generated. For demonstration, a 10-piece puzzle solution is loaded and details regarding the assembly software and how the program works is outlined.

Background

A probabilistic approach is taken which attempts to compute and maximize the joint probability of the fragment data given an assumed set of correspondences between the fragment boundaries and that the collection of all matched fragments share the same central axis and profile curve, i.e., the profile of the pot when viewed from the side is shared by all fragments when they are correctly aligned. The output of the system is a collection of solutions, listed in order of their probability.

Additionally, one can track the assembly process which starts with pairs of fragments that match well and then proceeds by merging a pair with another configuration of fragments, i.e., if fragment pairs (A,B) and (B,C) match well then the triplet (A,B,C) is a triplet formed by merging two pairs. The assembly process continues merging fragment groups (called configurations) until all of the fragments have been merged. At each stage, the algorithm merges a fragment group with a fragment pair that has the highest joint likelihood of being correct prior to computing the exact match.