SSD-DIS dataset

SSD-DIS dataset designed for the development and performance evaluation of document shadow removal algorithms. It consists of 12,224 groups of shadowed document images, which are composed of three parts: shadow-free original documents, shadow masks that delineate shadow regions, and artificially generated shadowed document images created by combining the original documents with the shadow masks. Given the large volume of shadow-free document data, both the dataset and the processing scripts are stored on Baidu Netdisk. Please use the following links to access them：Dataset: [Link：https://pan.baidu.com/s/5ZmiPhoGZz1kqI0KxFl93aQ]Processing Scripts: [Link：https://pan.baidu.com/s/5U0DKj6uKvBq_EAJHfOmPJg]Note: This dataset is for learning and research purposes only, not for commercial use.