A recursive analysis for form cell recognition

Abstract

It is very difficult to analyze form structures because of breaks in lines and ackhtional noises on the form image. This paper focuses on cell recognition in lowquality form images. The recognition method has two features to achieve robustness in cell recognition. One is grid representation using several types of intersection and the terminal points of the frame lines. The other is the recursive modification of the representation. A new representation is created according to the determination of the breaks in the line and the hypothesized location of the missed intersections by using the previous representation. The modification is processed recursively until the representation has perfect consistency and all form cells are detected In an experiment using 1565 form samples, all cells in 1538 samples (98.3% of 1565 samples) were recognized correctly by this method.

Recognition of Tables and Forms

“…A grid representation of the table/form can be also generated with the type of the intersections, the terminal points of the lines, and the imaginary lines [29]. Then, a recursive analysis is done to modify the grid to correct the cell detection.…”

Handbook of Document Image Processing and Recognition

“…For example, like in [30], cells (boxes) can be labeled first with four classes (box type), and this labeling should be done Fig. 19.6 Detection and modification of inconsistent grid-cross-points [29] automatically, but the paper does not explain how it could be done. A graph is built where a node is a box with a box type, and an edge is an adjacency of two boxes (Fig.…”

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

“…Among them, [14] and [12] use corners and junctions in the image to define the fields, [5] uses vertical and horizontal projections, and [1] employs special structures of connected strips to extract the field boundaries. Some recent works deal with special situation such as color images [10] or patterned backgrounds and noisy images [13], [9], [17].…”