Student Topics

Table Detection in Document Images

Status: open
Supervisors: Florian Kleber, Markus Diem

Problem Statement

Documents can contain tables, which can be placed everywhere within a document page, and a have a regular structure. The detection and analyis of tables is a part of current Document Image Analysis (DIA) systems. The detection of tables is of interest, since tables normally contain specific data.

The following image shows a document page, with figures and tables.

table

Goal

The goal of the Bachelor/Master thesis is the detection of tables within a document page, and the analysis of the structure. Afterwards the filled-in information of the table, and the structure should be available.

Workflow

  • Literature research
  • Implementation in Matlab or C++ with OpenCV
  • Development and evaluation of the system
  • Written report or bachelor thesis (in English) and final presentation

Requirements

  • Matlab, or C++ and OpenCV
  • Document Analysis VU