Dense-Haze
收藏资源简介:
Dense-Haze数据集是由瑞士苏黎世联邦理工学院计算机视觉实验室创建的,包含33对真实雾霾和对应的清晰户外场景图像。该数据集通过专业雾霾机生成浓密且均匀的雾霾场景,旨在推动单图像去雾技术的进步,特别是在处理真实且多样的浓雾场景方面。数据集的创建过程中,确保了所有户外场景在多云的早晨或傍晚进行拍摄,以保持光照条件的一致性。Dense-Haze数据集主要用于评估和提升现有单图像去雾技术的性能,特别是在处理浓雾条件下的图像清晰化问题。
The Dense-Haze dataset was created by the Computer Vision Laboratory at ETH Zurich, Switzerland. It contains 33 pairs of real hazy images and their corresponding clear outdoor scene images. This dataset generates dense and uniform hazy scenes using a professional haze generator, aiming to advance single-image dehazing technology, especially in addressing realistic and diverse dense fog scenarios. During the dataset creation process, all outdoor scenes were captured on cloudy mornings or evenings to maintain consistent lighting conditions. The Dense-Haze dataset is primarily used to evaluate and improve the performance of existing single-image dehazing technologies, particularly for image clarification tasks under dense fog conditions.

- Dense-Haze数据集首次发表,由Zhang等人提出,旨在解决图像去雾问题。
- Dense-Haze数据集首次应用于深度学习模型训练,显著提升了图像去雾算法的性能。
- Dense-Haze数据集被广泛应用于多个国际计算机视觉会议和竞赛中,成为图像去雾领域的重要基准数据集。
- Dense-Haze数据集的扩展版本发布,增加了更多复杂场景和不同天气条件下的图像数据,进一步推动了去雾技术的发展。



