Highway Vehicle Surveillance Dataset (HVSD)

Geo-Heinz Vehicle Dataset (GHVD) is a carefully selected collection of highway monitoring videos designed to support research in vehicle detection, tracking, and intelligent transportation systems. The dataset consists of original video clips captured under different environmental conditions—sunny, nighttime, and rainy—mimicking real-world dynamic highway traffic environments. In addition to the raw videos, GHVD includes the processed result from our new Geo-Heinz Mean Corner Detection Algorithm. The result contains annotated bounding boxes, vehicle IDs, and other corner detection information from the original videos. This dataset offers researchers a robust testbed for experimenting with real-time detection and tracking algorithms and allows comparative study with traditional and deep learning–based approaches, ultimately leading to increased reproducibility and follow-on innovations in intelligent transportation systems.

地理-海因茨车辆数据集（Geo-Heinz Vehicle Dataset，GHVD）是一套经过精心甄选的高速公路监控视频集，旨在为车辆检测、跟踪及智能交通系统领域的研究提供支撑。该数据集涵盖了在晴天、夜间、雨天等多种环境条件下采集的原始视频片段，可模拟真实世界中的动态高速公路交通场景。除原始视频外，GHVD还包含基于我们全新研发的地理-海因茨平均角点检测算法（Geo-Heinz Mean Corner Detection Algorithm）生成的处理结果，该结果包含原始视频中的标注边界框、车辆ID以及其他角点检测相关信息。本数据集为研究人员提供了稳健的测试平台，可用于开展实时检测与跟踪算法相关实验，同时支持与传统方法及基于深度学习的方法开展对比研究，最终提升智能交通系统领域研究的可复现性，并推动该领域后续的创新发展。