Hamlyn Rectified Dataset: Stereo Endoscopy Videos for Surgical Robotics

A rectified stereo video dataset from in vivo laparoscopic procedures, created by the Hamlyn Centre at Imperial College London. The repository contains rectified stereo images, calibration data, and ground truth, adapted for integration into the VSLAM-LAB framework. The dataset originates from the Endo-Depth-and-Motion project for 3D reconstruction and tracking in endoscopic videos.

Use Cases

• Training depth estimation models based on rectified stereo image pairs.
• Developing visual SLAM systems for endoscopic navigation using the provided ground truth.
• Benchmarking 3D reconstruction algorithms in surgical environments using the calibration data.
• Simulating robotic-assisted surgery scenarios with real in vivo video data.

Strengths

• Provides rectified stereo images, which are a processed form ready for computer vision tasks.
• Includes calibration data and ground truth, which are essential for quantitative evaluation.
• Originates from a known medical robotics research center, the Hamlyn Centre at Imperial College London.

Limitations

• Description metadata is limited; actual data quality, scale, and file formats require manual inspection after download.
• Row count and dataset size are unknown, which may limit suitability assessment for large-scale training.

Provenance

Source

Hamlyn Centre Laparoscopic at Imperial College London.

Collection Method

Recorded from in vivo endoscopy procedures; rectified and packaged for research.

Freshness

Last updated 2026-03-07 10:11:07; freshness should be verified.