3DTeethSeg'22 Dataset

该数据集包含了总共1800个口腔内3D扫描图像，来自900位患者，图像经过了牙齿定位、分割和标记的注释。提供了以JSON格式存储的真实标签和牙齿实例，并使用全球唯一标识符（uuid）对患者的信息进行了匿名处理。数据规模方面，训练数据中包含了900位患者的1800次扫描，共16004颗牙齿，而测试数据中则包含7995颗牙齿。该数据集的任务包括牙齿定位、分割和标记。

This dataset contains a total of 1800 intraoral 3D scanned images from 900 patients. The images have been annotated with tooth localization, segmentation and labeling. Ground truth labels and tooth instances stored in JSON format are provided, and patient information is anonymized using universally unique identifiers (UUIDs). In terms of data scale, the training subset includes 1800 scans from 900 patients, totaling 16,004 teeth, while the test subset contains 7,995 teeth. The tasks supported by this dataset include tooth localization, segmentation and labeling.

搜集汇总

背景与挑战

背景概述

3DTeethSeg'22 Dataset是一个包含1800个3D口腔内扫描的数据集，覆盖900名患者的上颌和下颌，用于牙齿分割和标记研究。数据集提供了牙齿标签和实例信息，旨在解决牙齿形状和位置变化带来的挑战。