Financial Table Extraction in Image Documents

TL;DR

This paper introduces an end-to-end deep learning pipeline for accurately extracting and transcribing financial tables from images, preserving spatial relationships, addressing a longstanding challenge in financial document analysis.

Contribution

It presents a novel integrated approach combining image segmentation, OCR, and sequence modeling for high-fidelity financial table extraction from images.

Findings

High accuracy in table detection and extraction

Preserves original spatial relations effectively

Applicable to various financial document formats

Abstract

Table extraction has long been a pervasive problem in financial services. This is more challenging in the image domain, where content is locked behind cumbersome pixel format. Luckily, advances in deep learning for image segmentation, OCR, and sequence modeling provides the necessary heavy lifting to achieve impressive results. This paper presents an end-to-end pipeline for identifying, extracting and transcribing tabular content in image documents, while retaining the original spatial relations with high fidelity.