Current Status and Performance Analysis of Table Recognition in Document Images With Deep Neural Networks

Khurram Azeem Hashmi,Muhammad Zeshan Afzal,Muhammad Ahtsham Afzal,Muhammad Adnan Afzal,Marcus Liwicki,Didier Stricker

doi:10.1109/access.2021.3087865

Abstract

The first phase of table recognition is to detect the tabular area in a document. Subsequently, the tabular structures are recognized in the second phase in order to extract information from the respective cells. Table detection and structural recognition are pivotal problems in the domain of table understanding. However, table analysis is a perplexing task due to the colossal amount of diversity and asymmetry in tables. Therefore, it is an active area of research in document image analysis. Recent advances in the computing capabilities of graphical processing units have enabled the deep neural networks to outperform traditional state-of-the-art machine learning methods. Table understanding has substantially benefited from the recent breakthroughs in deep neural networks. However, there has not been a consolidated description of the deep learning methods for table detection and table structure recognition. This review paper provides a thorough analysis of the modern methodologies that utilize deep neural networks. Moreover, it presents a comprehensive understanding of the current state-of-the-art and related challenges of table understanding in document images. The leading datasets and their intricacies have been elaborated along with the quantitative results. Furthermore, a brief overview is given regarding the promising directions that can further improve table analysis in document images.

Highlights

IntroductionTables are the prevalent means of representing and communicating structured data [1]
T ABLE understanding has gained an immense attraction since the last decade
Before throwing some light on the performance evaluation, it is appropriate to talk about the evaluation metrics first which are adopted to assess the performances of discussed approaches

Summary

Introduction

Tables are the prevalent means of representing and communicating structured data [1]. With the rise of Deep Neural Networks (DNN), various datasets for table detection, segmentation, and recognition have been published [2], [3]. This allows the researchers to employ the DNN to improve state-of-the-art results. One of the earlier works in the area of table analysis has been done by Kieninger et al [8]–[10]. Along with detecting the tabular area, their system known as T-Recs extracts the structural information of the tables

Methods

Results

Conclusion