Towards Automating Artifact Analysis: A Study Showing Potential Application in Computer Vision and Morphometrics to Artifact Typology
|Title||Towards Automating Artifact Analysis: A Study Showing Potential Application in Computer Vision and Morphometrics to Artifact Typology|
|Publication Type||Book Chapter|
|Year of Publication||2010|
|Authors||Lenardi, M, Merwin, DE|
|Book Title||Morphometrics for Nonmorphometricians|
|Keywords||computer vision, lithics typology, morphometrics, New York|
First released in 1961, William A. Ritchie’s A Typology and Nomenclature of New York State Projectile Points remains an influential resource for archaeologists working in the northeastern US. Ritchie examined a large and diverse sample of stone projectile points, but not all fit neatly into his typology. In this chapter, we propose that alternative approaches afforded by computer vision and morphometrics can shed light on this and other problems of traditional stone tool typology. With the advent of computer vision, we can now examine the entire morphological continuum of projectile points through statistical shape analysis. Following automated image capture, three analytical methods were evaluated using silhouette, outline, and landmark data. Biased free capture methods to record the form of an individual projectile point, together with using invariant shape descriptors to quantify the data, may result in more objective analysis than was possible in the past.