TNCR: Table net detection and classification dataset

We present TNCR, a new table dataset with varying image quality collected from open access websites. TNCR can be used for detection in scanned document images and their classification into 5 different classes. contains 9428 labeled tables approximately 6621 images. In this paper, we have implemented state-of-the-art deep learning-based methods to create several strong baselines. Deformable DERT Resnet-50 Backbone Network achieves the highest performance compared other precision of 86.7%, recall 89.6%, f1 score 88.1% on dataset. made source hope encouraging more learning approaches detection, structure recognition. The trained model checkpoints are available at https://github.com/abdoelsayed2016/TNCR_Dataset.