The corpus consists of three sets: 1-Shape-based features 2- Textual reference based features 3- Hybrid of shape based and textual reference based The corpus is capable of integrating plagiarism detection schemes by utilizing both the figures and their related text, both inside and outside the figure. The corpus will help out the research community to detect the plagiarism in the figures copied from different sources illegally. This will help the researchers to remain the sole proprietary of their contributions, i.e., figures makers.

Here are a few use cases for this project:

1. Academic Research Categorization: Researchers can use this model to scan and understand academic papers quickly. This would be particularly useful when working on a literature review, where hundreds of articles might be scanned to find relevant content.

2. Document Organization in Libraries: Libraries can use this model to help properly catalog and classify new books by scanning text to accurately identify the various metadata components.

3. Proofreading Tool for Publishers: The model could be used by publishing companies to check the structure of written content, ensuring all elements such as title, author, chapters, etc., are present and in the correct place.

4. Plagiarism Detection: Universities could use the model to help detect plagiarized works, as the model can extract the author, university, date, and other elements to cross-check databases.

5. Data Extraction and Analysis Tool: Companies can use the model to extract data from reports, documents, and whitepapers for further analysis. By recognizing elements like graphs, tables and figure captions, the model ensures all relevant information is captured.