RGB-T图像数据集
收藏资源简介:
RGB-T图像数据集由安徽大学创建,包含821对空间对齐的RGB-T图像及其用于显著性检测的地面实况注释。该数据集在不同场景和环境条件下记录,具有高多样性,并针对不同显著性检测算法进行了11种挑战性注释。数据集创建过程中考虑了显著对象的类别、大小、数量和空间信息,以增强多样性和挑战性。该数据集主要用于解决复杂场景中的图像显著性检测问题,通过集成RGB和热(RGB-T)数据,有效提升显著性检测性能。
The RGB-T image dataset, developed by Anhui University, comprises 821 pairs of spatially aligned RGB-T images along with their ground truth annotations for saliency detection. Captured across various scenarios and environmental conditions, this dataset features high diversity and includes 11 types of challenging annotations tailored for different saliency detection algorithms. During the dataset construction, factors including the categories, sizes, quantities, and spatial information of salient objects were taken into consideration to enhance its diversity and challenge level. This dataset is primarily intended to address the problem of image saliency detection in complex scenes, and it effectively improves the performance of saliency detection by integrating RGB and thermal (RGB-T) data.




