US 2019/0163971 A1Published application

Text line segmentation method

AssigneeKonica Minolta Laboratory U.S.A., Inc.

InventorsShubham Agarwal, Yongmian Zhang

Life of this patent

Estimating the term…

Loading patent…

Abstract

In a text line segmentation process, connected components (CCs) in document image are categorized into three subsets (normal, large, small) based on their sizes. The centroids of the normal size CCs are used to perform line detection using Hough transform. Among the detected candidate lines, those with line bounding box heights greater than a certain height are removed. For each normal size CC, if its bounding box does not overlap the bounting box of any line with an overlap area greater than a predefined fraction of the CC bounding box, a new line is added for this CC, which passes through the centroid of the CC and has an average slant angle. Each large size CCs are broken into two or more CCs. All CCs are then assigned to the nearest lines. A refinement method is also described, which can take any text line segmentation result and refine it.

Claims20

Dependent claims

A text line segmentation method performed on a binary document image, comprising:

(a) detecting connected components in the document image, and computing centroids and bounding boxes of the connected components;
(b) categorizing the connected components into three subsets based on their bounding box sizes, including a first subset of normal size connected components, a second subset of large size connected components, and a third subset of small size connected components;
(c) detecting a plurality of candidate lines from the first subset of connected components by applying Hough transform to the centroids of the first subset of connected components;
(d) removing candidate lines that has a line bounding box height greater than a predefined multiplier times an average line bounding box height of all candidate lines, wherein remaining candidate lines constitute legitimate lines;
(e) for each connected component of the first subset, computing an overlap area between the connected component and each legitimate line, and if a maximum one among the overlap areas is no greater than a predefined fraction of a bounding box area of the connected component, adding a new line as a legitimate line, wherein the new line passes through the centroid of the connected component and has an angle equal to an average angle of all candidate lines;
(f) breaking each of the second subset of connected components into two or more connected components;
and
(g) assigning each connected component of the first subset, each connected component of the third subset, and each connected component obtained by breaking the connected components of the second subset in step (f), to a nearest one of the legitimate lines obtained in step (e), to generate segmented text lines.

5 more elements

A computer program product comprising a computer usable non-transitory medium having a computer readable program code embedded therein for controlling a data processing apparatus, the computer readable program code configured to cause the data processing apparatus to execute a text line segmentation process on a binary document image, the process comprising:

(a) detecting connected components in the document image, and computing centroids and bounding boxes of the connected components;
(b) categorizing the connected components into three subsets based on their bounding box sizes, including a first subset of normal size connected components, a second subset of large size connected components, and a third subset of small size connected components;
(c) detecting a plurality of candidate lines from the first subset of connected components by applying Hough transform to the centroids of the first subset of connected components;
(d) removing candidate lines that has a line bounding box height greater than a predefined multiplier times an average line bounding box height of all candidate lines, wherein remaining candidate lines constitute legitimate lines;
(e) for each connected component of the first subset, computing an overlap area between the connected component and each legitimate line, and if a maximum one among the overlap areas is no greater than a predefined fraction of a bounding box area of the connected component, adding a new line as a legitimate line, wherein the new line passes through the centroid of the connected component and has an angle equal to an average angle of all candidate lines;
(f) breaking each of the second subset of connected components into two or more connected components;
and
(g) assigning each connected component of the first subset, each connected component of the third subset, and each connected component obtained by breaking the connected components of the second subset in step (f), to a nearest one of the legitimate lines obtained in step (e), to generate segmented text lines.

5 more elements

Dependencies are inferred from the claim text.

Facts

Application: 15/828,110Filed 30 Nov 2017
Published: 30 May 2019
Term ends: Not estimatedThe source records don’t support an estimate.

Classification12 codes

G06V 30/412
Analysis of document content: layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables
G06F 18/23213
Using statistics or function optimisation, e.g. modelling of probability density functions: with fixed number of clusters, e.g. K-means clustering
G06T 11/20
Two-dimensional [2D] image generation: drawing from basic elements
G06V 10/267
Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based…: by…