Repositori URV

Identificador: imarina:9330666

Handle: https://hdl.handle.net/20.500.11797/imarina9330666

Autors:
Bustos-Pérez, GOllé, A

Resum:
Lithic artifacts are some of the most common and numerous remains recovered from paleolithic archaeological sites. However, these materials can undergo multiple post-depositional alterations after their introduction into the archaeological record. Due to the high quantity of lithic remains recovered, a quick, flexible, and effective method for identifying degrees of alteration on the surface of lithic implements is highly desirable. The present study examines the use of gray level images to obtain quantitative data from the surface of flint artifacts and determine whether these images can detect the presence of post-depositional alterations. An experimental collection of flints was subjected to sequential episodes of rounding in a tumbling machine. After each episode, photographs were taken with a microscope, resulting in quantitative surface values using gray level values. The quantitative surface values were used as variables in machine learning models to determine time of exposure and the most salient variables for discrimination. Our results indicate that the extraction of metrics from gray level images successfully capture changes in the surface of flint artifacts caused by post-depositional processes. Additional results provide insight into which areas to sample when seeking post-depositional alterations and underscore the importance of particle size in the generation of alterations.