EndoScene
收藏资源简介:
EndoScene数据集是由巴塞罗那自治大学计算机视觉中心创建,旨在为结肠镜图像的端腔场景分割提供基准。该数据集包含912张图像,来源于44个视频序列,覆盖36名患者。数据集的创建过程中,结合了CVC-ColonDB和CVC-ClinicDB两个现有数据集,并增加了对腔体和镜面高光的标注。EndoScene数据集主要应用于结肠镜图像分析,特别是帮助临床医生在结肠镜检查中识别和分割息肉,以减少结直肠癌相关死亡率。
The EndoScene dataset was developed by the Computer Vision Center at the Autonomous University of Barcelona, aiming to provide a benchmark for endoluminal scene segmentation of colonoscopy images. It contains 912 images sourced from 44 video sequences covering 36 patients. During its creation, two existing datasets, CVC-ColonDB and CVC-ClinicDB, were combined, and annotations for body cavities and specular highlights were added. The EndoScene dataset is primarily applied to colonoscopy image analysis, particularly to help clinicians identify and segment polyps during colonoscopy so as to reduce mortality associated with colorectal cancer.

- 1A Benchmark for Endoluminal Scene Segmentation of Colonoscopy Images计算机视觉中心,计算机科学系,巴塞罗那自治大学,西班牙 · 2016年



