Assisting Creation Dating of 17th Century Dutch Drawings: Advances in Computational Art Connoisseurship Driven by Digital Image Processing

Rick Johnson with glasses and a white beard against a light background.

Event Date

Location
1127 Kemper Hall

Professor C. Richard Johnson

Cornell University

Abstract

Since 1900, the necessary requirement for using watermarks to assist in the dating of 17th-century artworks on laid paper was to confirm a wire-by-wire match of the watermarks of an artwork lacking an inscribed date and another sheet of paper with an inscribed date of use. The creation date of the undated artwork could be presumed to be close to the inscribed date of use of the other sheet. Creating physical overlays was difficult and essentially impractical across the large variety of watermarks in use in 17th-century Europe.

This talk describes a digital solution to this issue through the application of basic digital image processing. The bonus of this watermark overlay creation is that the underlying technique offers the first digital method for creating an image-searchable watermark library, which is necessary to reduce the number of possible matches from hundreds of choices to a limited few to be tested with creation of watermark overlays. Developing such an image-searchable watermark library has been a major goal since the advent of digital art history in the last decades of the twentieth century. This research is an indication of breakthroughs possible with the underexploited application of digital image processing to a basic task in a field seemingly outside the reach of digital computation.

Bio

Rick Johnson received the first PhD minor granted by Stanford University in Art History along with a PhD in Electrical Engineering in 1977. He is the Geoffrey S. M. Hedrick Senior Professor of Engineering Emeritus at Cornell University since 2021, after 40 years on the Cornell faculty. In 2007, Professor Johnson shifted his research from digital signal processing in control and communication systems to the application of digital image processing to the emerging field of computational art history. He has focused on the development of procedures for the matching of manufactured patterns in art supports, principally weave patterns in the canvases of Vincent van Gogh and Johannes Vermeer and watermarks in antique laid paper supports of the prints of Rembrandt van Rijn, the codices of Leonardo da Vinci, and, more recently, seventeenth-century Dutch drawings.

Event Category