Weather and Light Level Classification for Autonomous Driving Dataset

本数据集名为‘Weather and Light Level Classification for Autonomous Driving Dataset’，由德国的Valeo DAR Kronach和爱尔兰的Valeo 视觉系统共同创建。该数据集包含1165条图像数据，主要用于自动驾驶中的天气和光照水平分类。数据通过工业前视摄像头记录，分辨率为1024×1084，涵盖了多种天气和光照条件，如晴朗、中度光、低光、雾、雨和雪。数据集的创建过程涉及视频序列的录制和图像的抽样，并通过主动学习框架减少冗余。该数据集主要应用于自动驾驶系统中的视觉感知增强，特别是在恶劣天气和低光环境下的性能提升。

This dataset is named 'Weather and Light Level Classification for Autonomous Driving Dataset', and was co-created by Valeo DAR Kronach from Germany and Valeo Vision Systems from Ireland. It contains 1165 image samples, mainly intended for weather and light level classification tasks in autonomous driving scenarios. The data was captured using industrial forward-looking cameras with a resolution of 1024×1084, covering diverse weather and illumination conditions including clear sky, moderate light, low light, fog, rain and snow. The dataset construction process involves recording video sequences, sampling images and reducing redundant data via an active learning framework. This dataset is primarily applied to enhance visual perception in autonomous driving systems, especially to improve system performance under adverse weather and low-light environments.