A Deep Learning Approach to Assist in Pottery Reconstruction From Its Sherds

Pottery is one of the most common and abundant types of human remains found in archaeological contexts. The analysis of archaeological pottery involves the reconstruction of pottery vessels from their sherds, which represents a laborious and repetitive task. In this work, we investigate a deep learning-based approach for making that process more efficient, accurate and faster. In that regard, given a sherd’s digital point cloud in a standard, so-called canonical position, the proposed method predicts the geometric transformation which moves the sherd to its expected position relative to the respective vessel’s coordinate system. Among the main components of the proposed method, a pair of deep 1D-convolutional neural networks trained to predict the 3D Euclidean transformation parameters stands out. Herein, rotation and translation components are treated as independent problems, so, while the first network is dedicated to predict translation moments, the other infers the rotation parameters. In practical applications, once a vessel’s shape is identified, the networks can be trained for predicting the target transformation parameter values. Given the 3D model of a vessel, it is broken virtually countless times for the production of training data, which consist of a large set of virtual sherds. The herein proposed 1D-convolutional neural network architecture, so-called PotNet, was inspired by the PointNet architecture. While PointNet was motivated by 3D point clouds classification and segmentation applications, PotNet was designed to perform non-linear regressions. Experiments using three distinct real vessels were carried out, and the reported results suggest that the proposed method can be successfully used for aiding pottery reconstruction.