RGBT Crowd Counting (RGBT-CC)
收藏资源简介:
RGBT Crowd Counting (RGBT-CC)数据集由中山大学计算机科学与工程学院创建,包含2030对RGB-热图像,总计138,389个标注人物。该数据集通过光学和热学信息的结合,旨在提升不受约束场景中行人的识别能力。数据集涵盖多种场景,如商场、街道、火车站等,且在不同光照条件下(如白天和夜晚)采集。RGBT-CC数据集的应用领域包括城市人群分析,特别是在交通管理和视频监控中的潜在应用,以及在COVID-19疫情期间的社会距离监控。
The RGBT Crowd Counting (RGBT-CC) dataset was created by the School of Computer Science and Engineering, Sun Yat-sen University. It contains 2030 pairs of RGB-thermal images, with a total of 138,389 annotated pedestrians. By integrating optical and thermal modalities, this dataset aims to enhance pedestrian detection performance in unconstrained scenarios. The dataset covers diverse scenarios including shopping malls, streets, railway stations and more, and was collected under various lighting conditions such as daytime and nighttime. Application fields of the RGBT-CC dataset include urban crowd analysis, with potential applications in traffic management, video surveillance, as well as social distance monitoring during the COVID-19 pandemic.




