Template identification and matching for data analysis in large sets of documents

A system comprising:

• at least one memory configured to store computer-readable instructions;
• and
• at least one processor communicatively coupled with the at least one memory, and configured
• to execute the computer-readable instructions to:
• access a plurality of documents stored in one or more databases communicatively coupled with the at least one processor;
• convert each page of each document of the plurality of documents into a respective image that comprises a plurality of pixel key points and corresponding descriptors;
• identify, based upon the plurality of pixel key points and corresponding descriptors for each page of each document of the plurality of documents, a plurality of unique templates corresponding to each document of the plurality of documents;
• and
• batch a first list of documents based upon matching a first unique template of the plurality of unique templates included in a first document of the plurality of documents with other documents of the plurality of documents,
• wherein to identify the plurality of unique templates corresponding to each document of the plurality of documents, the at least one processor is configured to execute the computer-readable instructions to:
• compare, using a brute force k-nearest neighbors algorithm, the plurality of pixel key points and corresponding descriptors for each page of the first document against a plurality of pixel key points and corresponding descriptors for one or more other pages of the first document;
• based on the comparing, generate a plurality of distances identifying similarity among all pages of the first document;
• in accordance with the plurality of distances, obtain an image homography space and inliers and outliers of the plurality of distances corresponding to all pages of the first document;
• and
• based upon the image homography space and the inliers and outliers of the plurality of distances, identify the plurality of unique templates corresponding to the first document.

One or more non-transitory computer-readable media comprising machine-readable instructions stored thereon, which, when executed by at least one processor of at least one computing device communicatively coupled with one or more databases, cause the at least one processor to:

• access a plurality of documents stored in the one or more databases;
• convert each page of each document of the plurality of documents into a respective image that comprises a plurality of pixel key points and corresponding descriptors;
• identify, based upon the plurality of pixel key points and corresponding descriptors for each page of each document of the plurality of documents, a plurality of unique templates corresponding to each document of the plurality of documents;
• and
• batch a first list of documents based upon matching a first unique template of the plurality of unique templates included in a first document of the plurality of documents with other documents of the plurality of documents,
• wherein to identify the plurality of unique templates corresponding to each document of the plurality of documents, the at least one processor is configured to execute the computer-readable instructions to:
• compare, using a brute force k-nearest neighbors algorithm, the plurality of pixel key points and corresponding descriptors for each page of the first document against a plurality of pixel key points and corresponding descriptors for one or more other pages of the first document;
• based on the comparing, generate a plurality of distances identifying similarity among all pages of the first document;
• in accordance with the plurality of distances, obtain an image homography space and inliers and outliers of the plurality of distances corresponding to all pages of the first document;
• and
• based upon the image homography space and the inliers and outliers of the plurality of distances, identify the plurality of unique templates corresponding to the first document.