Detecting the Past: Advancements in Comic Panel Detection for Cultural Heritage Preservation

The preservation and restoration of historical murals is a crucial task in the field of cultural heritage preservation. In this study, we explored the use of comic panel detection technology for the preservation of historical murals. Specifically, we applied the faster Region-based Convolutional Neural Network (R-CNN) algorithm, a state-of-the-art object detection algorithm, for the detection and segmentation of individual panels in historical murals. We developed a dataset of high-resolution photographs of historical murals and evaluated the performance of the algorithm using several metrics, including precision, recall, F1 score, and mean average precision (mAP). The algorithm achieved high precision, recall, and F1 score values across all five folds, with an average mAP value of 0.86. The results demonstrate the effectiveness of the comic panel detection algorithm in accurately detecting and segmenting individual panels in historical murals. Additionally, we conducted a case study on a partially damaged historical mural, where the panel segmentation provided valuable information for the restoration process. The restoration process successfully restored the damaged areas of the mural while preserving the original composition and color scheme. Overall, this study highlights the potential of comic panel detection technology in the field of cultural heritage preservation and underscores the importance of technology in supporting these efforts.