Temporal Color Constancy (TCC) dataset
收藏资源简介:
Temporal Color Constancy (TCC)数据集是由坦佩雷大学的学生在自由时间内使用华为Mate 20 Pro手机收集的,包含600个不同长度的真实世界序列,覆盖室内外多种天气和光照条件,是目前最大的真实时序数据集。数据集的创建过程涉及使用特定的手机设置自动存储原始传感器图像,并通过日期和时间标签与拍摄帧关联。该数据集主要用于评估时序颜色恒常性方法,旨在解决传统单帧颜色恒常性方法的局限性,通过多帧序列提高场景光照颜色的估计准确性。
The Temporal Color Constancy (TCC) dataset was collected by students from Tampere University during their free time using the Huawei Mate 20 Pro smartphone. It contains 600 real-world sequences of varying lengths, covering diverse indoor and outdoor weather and lighting conditions, making it the largest real-world temporal dataset to date. The dataset creation process involved automatically storing raw sensor images with specific smartphone settings and associating each captured frame with its corresponding date and time stamp. This dataset is primarily used to evaluate temporal color constancy methods, aiming to address the limitations of traditional single-frame color constancy approaches and improve the estimation accuracy of scene illumination color via multi-frame sequences.




