BihoT
收藏资源简介:
BihoT数据集是由北京理工大学信息与电子学院和北京分形信号与系统重点实验室创建的大规模高光谱伪装目标跟踪数据集。该数据集包含41,912张高光谱图像,覆盖49个视频序列,主要用于解决在复杂环境中通过光谱信息进行目标跟踪的问题。数据集的创建过程中,收集了多种人工伪装场景下的图像,确保了数据集的多样性和挑战性。BihoT数据集的应用领域主要集中在高光谱伪装目标跟踪,旨在通过光谱特征提高跟踪算法的鲁棒性和准确性。
The BihoT dataset is a large-scale hyperspectral camouflaged object tracking dataset developed by the School of Information and Electronics, Beijing Institute of Technology, and the Beijing Key Laboratory of Fractal Signal and System. This dataset includes 41,912 hyperspectral images across 49 video sequences, and is primarily designed to address the problem of object tracking using spectral information in complex environments. During the construction of the dataset, images from various artificial camouflage scenarios were collected to ensure its diversity and challenging characteristics. The application scope of the BihoT dataset is mainly focused on hyperspectral camouflaged object tracking, aiming to improve the robustness and accuracy of tracking algorithms via spectral features.

- 1BihoT: A Large-Scale Dataset and Benchmark for Hyperspectral Camouflaged Object Tracking北京理工大学信息与电子学院,北京分形信号与系统重点实验室 · 2024年



