Answers are grounded in this paper and cite their sources.
Ancient stone tools excavated from ruins are important materials for archaeological research. The manufacturing process of stone tools is revealed by frequently assembling and disassembling the joining materials. Each stone tool is stored in a bag with its identification number and photo. However, storage and management of stone tools may occur human errors, such as mis-recognition and mistaking of the storage bag. This study proposes an image-based identification method to match each stone tool with its corresponding storage bag. Two commonly used stable postures were defined for each stone tool. The collected image dataset is used for deep learning of an untrained CNN model and seven pre-trained models. The experimental results showed better accuracy and processing speed than those of previous research. Finally, three of the models with high classification performances are selected to construct the detector with the YOLO framework for practical scenario.
No takes yet. Share an insight, caveat, or question.
You et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: