ECE 4210 home page

ECE 4210 / Fall 2015

ART SUPPORT ANALYSIS ALGORITHMS

Basic Info

Professor: Rick Johnson (email to johnson at ece.cornell.edu)

Class meeting time and place: Monday and Wednesday 8:40 - 9:55 am in 368 Hollister Hall and some Fridays (August 28 and September 4, 11, and 18) 8:40 - 9:55am in 362 Hollister Hall; All team meetings in 390 Rhodes Hall.

Course Website: within http://blackboard.cornell.edu

Catalog Description: (Prerequisite: ECE 3250) The analysis of fine art by art historians and conservators involves the close examination of various images of the works of art. The image data available for this course include digitized x-radiographs of paintings by Vincent van Gogh and Johannes Vermeer and beta-radiographs of prints by Rembrandt van Rijn. This course will provide a basic description of traditional manual examination tasks and of the building blocks of image processing that can be used to automate them. Students will employ, assess, and improve "starter" schemes performing canvas thread counting and thread angle measurement and laid paper chain line pattern matching. Matlab and Mathematica will be exploited in providing the computer code for the "starter" schemes and the presentation of image processing concepts.

Tentative Lecture Schedule

8/26: Course introduction

8/28: Signal processing in computational art history

8/31: Canvas x-radiograph and laid paper beta-radiograph datasets

9/2, 4: Image processing basics

9/7: Labor day holiday

9/9, 11, 14: Image processing basics (continued)

9/16: Algorithm building with synthesized data: Line spacing fat/thin detector (FTD) / Discussion of Tasks 1 and 2

9/18: Team progress report meetings (about dataset of laid paper radiographs)

9/21: FTD / Reports on tasks 1 and 2 and discussion of tasks 3 and 4

9/23: Team progress report meetings (about dataset of canvas radiographs)

9/28: FTD / Reports on tasks 3 and 4 and discussion of task 5

9/30: FTD / Reports on task 5

10/5: Reports on laid paper dataset examination

10/7: Reports on canvas dataset examination

10/12: Fall recess

10/14: No class

10/19, 21: Pulp to beta-radiograph

10/26: Chain-line pattern matching

10/28: Automated chain line extraction

11/2: Team progress report meetings (about automating determination of chain line spacing average)

11/4, 9: Fabric to x-radiograph

11/11, 16: Thread counting and weave matching

11/18: Angle maps, cusping, and weft snakes

11/23: Team progress report meetings (about automating fat/thin thread detector)

11/25: Thanksgiving recess

11/30: Reports on chain line spacing average extraction

12/2: Reports on fat/thin thread detection

Reading Material

Application reference material to be on reserve at Cornell Fine Arts Library

A. Kirsh and R. S. Levenson, Seeing Through Paintings: Physical Examination in Art Historical Studies, pp. 28-53, Yale University Press, 2000.
W. Liedtke, C. R. Johnson, Jr., and D. H. Johnson, ``Canvas Matches in Vermeer: A Case Study in the Computer Analysis of Fabric Supports,'' Metropolitan Museum Journal, 2012.
K. H. Lister, C. Peres, I. Fiedler, "Tracing an Interaction: Supporting Evidence, Experimental Grounds" in Van Gogh and Gauguin: The Studio of the South, ed. D. W. Druick and P. K. Zegers, pp. 354-369, Thames and Hudson, 2001.
L. van Tilborgh, T. Meedendorp, E. Hendriks, D. H. Johnson, C. R. Johnson, Jr., and R. G. Erdmann, ``Weave Matching and Dating of Van Gogh's Paintings: An Interdisciplinary Approach,'' The Burlington Magazine, vol. 154, pp. 112-122, February 2012.
E. van de Wetering, ``The Canvas Support'' (Chapter 5), Rembrandt: The Painter at Work, University of California Press, pp. 90-130, 2000.
E. van de Wetering, ``Canvas Research with Emil Bosshard, Remarks on Method,'' in Emil Bosshard, Paintings Conservator (1945-2006): Essays by Friends and Colleagues}, ed. M. De Peverelli, M. Grassi, H.-C. von Imhoff, Centro Di, pp. 256-269, 2009.
C. A. Baker, From the Hand to the Machine: Nineteenth-century American paper and mediums: technologies, materials, and conservation, Legacy Press, pp. 34-50, 2010.
E. Hinterding, Rembrandt as an Etcher: The practice of production and distribution, vol. 1, pp.15-65, Sound and Vision, 2006.
D. Kushel, ``Radiographic Methods Used in the Recording of Structure and Watermarks in Historic Paper'' in Fresh Woods and Pastures New: Seventeenth Century Dutch Landscape Drawings from the Peck Collection, Chapel Hill, pp. 116-123, 1999.

Techncial reference material to be on Cornell Library reserve

R. C. Gonzalez and R. E. Woods, Digital Image Processing (3rd ed.), Prentice Hall, 2007.
J. H. McClellan, R. W. Schafer, and M. A. Yoder, Signal Processing First, Pearson Prentice Hall, 2003.

Other publications relevant to course material available on http://people.ece.cornell.edu/johnson.

Recent special issues devoted to image processing for art investigation

IEEE Signal Processing Magazine vol. 25, no. 4, July 2008 and vol. 32, no. 4, July 2015
Signal Processing (EURASIP), vol. 93, issue 3, March 2013.

Grading

No exams.

The assignments in the course will be a mixture of individual and team assignments. Teams are to be composed of 2 or 3 (no more, no less) students. Each student will be on two teams: one focusing on laid paper chain line patterns and the other on canvas thread patterns. No pair of students can be on the same two teams. Each team will present two reports, each of which is worth 20% of the course grade. Homework assignments to be done individually will account for half of the course grade. The remaining 10% of the final grade is based on attendance, classroom participation, and office hour interaction.

Every student must turn in a report (in electronic and hard copy) for each individual homework assignment. Students are encouraged to discuss these ``individual'' problems and their solutions with each other. However, no code is to be exchanged on these ``individual'' problems and all prose included in the report is to be written by the individual submitting the report. Provide the name and netid on the cover sheet of your report of the students with whom you discussed that individual assignment.

Only one version of each team report is to be submitted (in electronic and hard copy). Electronic copies of reports can be emailed to johnson@ece.cornell.edu. An electronic copy can also be deposited in person with Professor Johnson's administrative assistant Daniel Richter in 314 Rhodes Hall. Hard copies are to be delivered in class to Professor Johnson or at other times to Daniel Richter. If delivered to Daniel, please be certain that he writes the time and date of submission on the report.

All reports are due in electronic and hard copy by noon on noted due date. Late report penalty: 10% of maximum score up to 24 hours late, 20% if late 24 to 48 hours, and not accepted for grading if over 48 hours late. No reports will be accepted for grading after noon on Thursday, December 3, 2015.

Homework Due Dates:

Note: Homeworks 1, 2, and 3 are individual assignments to be completed by fall recess. Homeworks 4 and 5 are team assignments that extend over the full term..

Schedule of Team Meetings (in Prof. Johnson's office) and Oral Reports (in class):

9/16, 21: Team progress report meetings - Agenda: confirmation and combination of data collected for Homework 1
9/30, 10/5: Team reports on findings regarding combined dataset
11/2: Team progress report meeting - Agenda: Status of Homework 4 results
11/23: Team progress report meeting - Agenda: Status of Homework 5 results
11/30, 12/2: Team reports on Homeworks 4 and 5

Each student in this course is expected to abide by the Cornell University Code of Academic Integrity. Any work submitted by a student in this course for academic credit will be the student's own work.

Last Revised August 19, 2015