Automated Template Based Signature Detection and Extraction from Multilingual Documents

In this work, the model proposes a method for detecting signatures from multilingual documents. The model has explored the templatebased matching techniques such as Normalized Cross Correlation (NCC) and morphological processing segmentation to enhance the effectiveness of detection and extraction of signatures from the scanned documents. The model uses a matching technique to establish and study the similarity between the extracted and template signatures. The use of template-based techniques provides a solid foundation for successfully detecting and extracting signatures across the different languages. The proposed method extracts the signatures with an overall accuracy of $\mathbf{9 4. 4 \%}$ for the dataset considered. The methodology’s excellent accuracy demonstrates its dependability and applicability in real-world scenarios. As there is a demand for quick and automatic processing of documents for signature identification or verification, the method offers promising results and ensures digital document legitimacy. The methodology represents a significant advancement in signature detection and extraction from multilingual documents.