Automated Monitoring of Cultural Heritage Artifacts Using Semantic Segmentation
This addresses the problem of efficient preservation for cultural heritage managers, but it is incremental as it applies existing methods to a new domain.
The paper tackled automated crack detection for cultural heritage preservation by comparing U-Net architectures with different CNN encoders for semantic segmentation, showing promising generalization to unseen statues and monuments without explicit training on such images.
This paper addresses the critical need for automated crack detection in the preservation of cultural heritage through semantic segmentation. We present a comparative study of U-Net architectures, using various convolutional neural network (CNN) encoders, for pixel-level crack identification on statues and monuments. A comparative quantitative evaluation is performed on the test set of the OmniCrack30k dataset [1] using popular segmentation metrics including Mean Intersection over Union (mIoU), Dice coefficient, and Jaccard index. This is complemented by an out-of-distribution qualitative evaluation on an unlabeled test set of real-world cracked statues and monuments. Our findings provide valuable insights into the capabilities of different CNN- based encoders for fine-grained crack segmentation. We show that the models exhibit promising generalization capabilities to unseen cultural heritage contexts, despite never having been explicitly trained on images of statues or monuments.