Image-based Archaeological Object Detection Using Machine Learning

EasyChair Preprint 6667

Abstract

This paper presents the image dataset of archeological objects of different categories which are found in Bangladesh. This dataset contains 24 number of statues named Aghora,Vishu, Uma-Mahesvara, Kalayana Sundara,  Mahapratisara, Vishnu (Narasimha Avatara), Surya, Ganesha, Mahamaya, Gauri, Durga, Gajalaksmi, Manjuvara, Avalokitesvara, Chamunda, Akshobhya (Tathagata Buddha), Nilkhantha, Garuda, Sivalinga, Manjusri, Votiva Stupa, Ventilator, Lintel Yaksha. The dataset also contains 4 types of all stone named Black Basalt, Black Stone, Sand Stone, Granite Stone. As Bangladesh Natinional Museum is the national museum of the country,so we have gotten varios kinds of archeological statues here. These statues are the historical view of the country. People can find the culture which were used to follow by the people long time ago. People can use this paper for gathering knowledge, getting information for experiments. In our dataset, all the pictures are captured by the camera of smart phone having enough image pixel information for applying image processing tools. We have created an image dataset consisting of 39385 sample images. The dataset can be considered as the basic ground-truth dataset for building any mobile application based on archeological object detection.

Keyphrases: Statue classification, archeology, image processing, machine learning, stone type

@booklet{EasyChair:6667,
  author    = {Sarmistha Pal Trisha and Hena Akhter and Md. Sabbir Hossain and Rubaiya Hafiz},
  title     = {Image-based Archaeological Object Detection Using Machine Learning},
  howpublished = {EasyChair Preprint 6667},
  year      = {EasyChair, 2021}}