Document layout structure extraction using bounding boxes of different entitles

Abstract

This paper presents an eficient technique for document page layout structure extraction and classification by analyzing the spatial configuration of the bounding boxes of different entzties on the given image. The algorithm segments an image into a list of homogeneous zones. The classification algorithm labels each zone as text, table, line-drawing, halftone, ruling, or noise. The text-lines and words are extracted within text zones and neighboring text-lines are merged to form text-blocks. The tabular structure is further decomposed into row and column items. Finally, the document layout hierarchy is produced from these extracted entities.

Performance Evaluation of Document Structure Extraction Algorithms

“…In this section, we briefly present a rule-based algorithm that extracts document layout structure using the bounding boxes of different entities [15]. Then we report the performance of each module on the images from the UW-III Document Image Database.…”

A Model-Based Field Frame Detection for Handwritten Filled-in Forms

“…Since then, many other works dealing with high-level form representation, studying the structural relation among fields and often pursuing a completely automatic form analysis have been presented. Some of them, are essentially rule-based, as [7], [6], [16] and [18]. Other use graphs to establish relations among the fields, as [2] and [20].…”

Hybrid page segmentation using multilevel homogeneity structure

“…These methods typically separate the original document into many different regions. Then use many filters to classify each region [5,18,20] (only one level of homogeneous region is used). In addition to creating many filters, these methods only effective when the region is not too complicated.…”