MITI Dataset

MITI数据集是由慕尼黑工业大学外科手术MITI研究组创建，用于评估腹腔镜手术中的立体视觉惯性计算机视觉算法。该数据集包含从手持手术干预中完整记录的多模态传感器信息，包括IMU、立体视频和红外跟踪数据，用于评估SLAM/SfM/3D重建/视觉惯性里程计算法。数据集选择了一个包含极少切割和组织变形的手术案例，以全扫描腹部的方式，手持相机进行记录，非常适合测试SLAM算法。该数据集旨在推动适用于MIS应用的视觉惯性算法的发展，帮助研究人员提升算法性能。

The MITI Dataset was created by the MITI Surgical Research Group at the Technical University of Munich, for evaluating stereo visual-inertial computer vision algorithms in laparoscopic surgery. This dataset comprises complete multimodal sensor data recorded during handheld surgical interventions, including IMU, stereo video, and infrared tracking data, which is designed to assess SLAM, SfM, 3D reconstruction, and visual-inertial odometry algorithms. The dataset is based on a surgical case with minimal tissue cutting and deformation, recorded using a handheld camera conducting full abdominal scanning, making it highly suitable for testing SLAM algorithms. This dataset aims to advance the development of visual-inertial algorithms tailored for minimally invasive surgery (MIS) applications, and support researchers in enhancing their algorithm performance.