SERV-CT

displayName: 'SERV-CT (SERV-CT: A disparity dataset from CT for validation of endoscopic
3D reconstruction)'
license:
- CC BY 4.0
mediaTypes:
- Image
paperUrl: https://arxiv.org/pdf/2012.11779v1.pdf
publishDate: "2020"
publishUrl: https://www.ucl.ac.uk/interventional-surgical-sciences/serv-ct
publisher:
- University of London
- National Center for Tumor Diseases
- German Cancer Research Center
tags:
- Endoscope
taskTypes:
- 3D Reconstruction

---
# 数据集介绍
## 简介
手术场景的内窥镜立体重建会产生特定的问题，包括缺乏清晰的角落特征、高度镜面反射的表面特性以及血液和烟雾的存在。这些问题给立体重建本身和标准化数据集生产带来了困难。我们提出了一个基于锥形束 CT (SERV-CT) 的立体内窥镜重建验证数据集。在内窥镜视野内放置两具离体小猪全躯干尸体，在 CT 扫描中可以看到内窥镜和目标解剖结构。手动对齐内窥镜的后续方向以匹配立体视图，并计算基准差异、深度和遮挡。 CT 扫描的要求将每个离体样本的立体对数限制为 8 个。对于第二个样本，获取 RGB 表面以帮助对齐平滑、无特征的表面。重复的手动对齐显示了大约 2 个像素的 RMS 视差精度和大约 2 mm 的深度精度。提供了一个简化的参考数据集，由内窥镜图像对组成，具有相应的校准、差异、深度和遮挡，涵盖了大部分内窥镜图像和一系列组织类型，包括光滑的镜面，以及深度的显着变化。 SERV-CT 数据集为手术应用提供了易于使用的立体验证，具有平滑的参考差异和深度，覆盖了大部分内窥镜图像。
## 引文
```
@article{edwards2020serv,
title={SERV-CT: A disparity dataset from CT for validation of endoscopic 3D reconstruction},
author={Edwards, PJ and Psychogyios, Dimitris and Speidel, Stefanie and Maier-Hein, Lena and Stoyanov, Danail and others},
journal={arXiv preprint arXiv:2012.11779},
year={2020}
}
```

## Download dataset
:modelscope-code[]{type="git"}

displayName: 'SERV-CT (SERV-CT: A disparity dataset from CT for validation of endoscopic 3D reconstruction)'
license:
- CC BY 4.0
mediaTypes:
- Image
paperUrl: https://arxiv.org/pdf/2012.11779v1.pdf
publishDate: "2020"
publishUrl: https://www.ucl.ac.uk/interventional-surgical-sciences/serv-ct
publisher:
- University of London
- National Center for Tumor Diseases
- German Cancer Research Center
tags:
- Endoscope
taskTypes:
- 3D Reconstruction

---
# Dataset Introduction
## Overview
Endoscopic 3D reconstruction in surgical scenarios poses specific challenges, including the lack of distinct corner features, highly specular reflective surface properties, and the presence of blood and smoke. These issues introduce difficulties for both 3D reconstruction itself and the production of standardized datasets. We propose SERV-CT, a stereo endoscopic reconstruction validation dataset based on cone-beam CT. Two ex vivo whole piglet cadavers were placed within the endoscopic field of view, with the endoscope and target anatomical structures visible in the CT scans. The subsequent orientations of the endoscope were manually aligned to match stereo views, and ground-truth disparity, depth, and occlusion were calculated. The constraints of CT scanning limited the number of stereo pairs per ex vivo specimen to 8. For the second specimen, RGB surface scans were acquired to aid alignment of smooth, featureless surfaces. Repeated manual alignments demonstrated a RMS disparity accuracy of approximately 2 pixels and a depth accuracy of around 2 mm. A simplified reference dataset is provided, consisting of endoscope image pairs with corresponding calibrations, disparity maps, depth maps, and occlusion masks, covering most endoscopic images and a range of tissue types including smooth specular surfaces, as well as significant variations in depth. The SERV-CT dataset provides an easy-to-use stereo validation tool for surgical applications, with smooth ground-truth disparity and depth maps covering most endoscopic images.
## Citation

@article{edwards2020serv,
title={SERV-CT: A disparity dataset from CT for validation of endoscopic 3D reconstruction},
author={Edwards, PJ and Psychogyios, Dimitris and Speidel, Stefanie and Maier-Hein, Lena and Stoyanov, Danail and others},
journal={arXiv preprint arXiv:2012.11779},
year={2020}
}

## Download dataset
:modelscope-code[]{type="git"}

搜集汇总

背景与挑战

背景概述

SERV-CT是一个基于锥束CT的内窥镜立体重建验证数据集，旨在解决手术场景中因缺乏清晰特征、高镜面反射和干扰物（如血液和烟雾）带来的重建挑战。数据集包含两个离体猪尸体的内窥镜图像对，每对图像配有校准、视差、深度和遮挡信息，覆盖多种组织类型和深度变化，提供约2像素视差精度和2毫米深度精度的平滑参考数据。