GMRP数据集

GMRP数据集是专为地面移动机器人设计的，包含3896对图像及其用于可行驶区域和道路异常的地面实况。该数据集由香港科技大学电子与计算机工程系创建，旨在为地面移动机器人提供一个评估和改进其环境感知能力的基准。数据集内容包括多种视觉特征，如RGB图像、视差图像等，用于训练和测试深度学习模型。创建过程中，研究者利用了先前发表的地面移动机器人感知数据集，并在此基础上进行了扩展和优化。该数据集主要应用于地面移动机器人的可行驶区域和道路异常检测，旨在提高机器人在复杂环境中的自主导航能力和安全性。

The GMRP dataset is specifically designed for ground mobile robots, comprising 3896 pairs of images and their corresponding ground truths for drivable area and road anomaly detection. Developed by the Department of Electronic and Computer Engineering at The Hong Kong University of Science and Technology, this dataset serves as a benchmark for ground mobile robots to evaluate and improve their environmental perception capabilities. The dataset includes various visual modalities such as RGB images, disparity images, and others, which are used for training and testing deep learning models. During its creation, researchers utilized previously published ground mobile robot perception datasets and expanded and optimized them to build this dataset. This dataset is mainly applied to drivable area and road anomaly detection for ground mobile robots, with the goal of enhancing the autonomous navigation capability and safety of robots in complex environments.