Skip to content

ascota_classification

ascota_classification is the second stage of the ASCOTA pipeline. Currently in development.

Purpose

In this package, we focus on:

  • Classification of sherd types: Using deep learning models to classify pottery fragments into predefined categories based on shape and size.

  • Color Classification: Extracting relevant color features from segmented and color-corrected images to support classification and analysis.

  • Texture Classification: Identifying and labeling surface textures of pottery fragments to aid in classification and comparative analysis.

  • Pattern/Decoration Classification: Use vision transformer models to identify and classify decorative patterns on pottery fragments.

Submodules

  • color: color-based classification of pottery sherds.
  • decoration: decoration pattern classification of pottery sherds.
  • type: type classification of pottery sherds.