EP0496531A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Method and system for layout analysis of a document image
Assignee: International Business Machines Corp
•
Filing: 1992-01-16
•
Granted: —
•
Expires: 2012-01-16
Abstract
Abstract An object of the invention is to precisely detect objects from a document image in layout analysis on the basis of a model with minimal information given to the model. Another object is to sufficiently cope with changes in position and size of the objects with a single model. A model contains data on spatial relationships of objects, data indicative for each of the objects of whether its existence on a page is mandatory or not, data for each of the objects corresponding to terminal nodes on the number of included character strings, and data for each of the objects corresponding to intermediate nodes on the number of immediate successors. From a document image are extracted not only character string regions 1 to 7 but also separators segmenting the image into objects and sub-separators treated as candidate boundaries. Labels indicative of object names are assigned to the character strings 1 to 7 by using the separators and sub-separators as constraints.
Inventors
- 👤 Akio Yamashita
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-30 22:35
CPC Classifications
H; H04; H04N; H04N1/00; H04N1/40; H04N1/40062; G; G06; G06V; G06V30/00; G06V30/40; G06V30/41; G06V30/416
Claims Analysis
Total Claims: 10
Claim 1
Independent Claim
95 words
A method for the layout analysis of an input document image comprised of objects arranged in a hierarchical tree structure, the method comprising:- (a) registering a layout model representing spatial relations of objects in a model document; (b) extracting character string regions from the input document image; (c) detecting separators and generating data on their positions, said separators being white and/or black regions satisfying predetermined conditions; (d) segmenting said document image into tree-structured sub-areas by using said separator data and (e) matching said character string regions with said layout model by using separators as boundaries.
Claim 2
Dependent Claim
52 words
The method of claim 1 further comprising the steps of detecting sub-separators within each of said sub-areas and generating data on their positions, said sub-separators being white regions satisfying predetermined conditions different from said separators, the matching of character strings with the layout model employing the sub-separators as boundary candidates for objects.
Claim 3
Dependent Claim
51 words
The method according to claim 2 including the steps of generating data on character size for each character string region, detecting a character string region that is different in character size from the immediately precedent character string region, and detecting a white region above the detected character string region as sub-separator.
Claim 4
Dependent Claim
42 words
The method according to claim 2 or 3 including the steps of generating pitch data for each of white regions between vertically successive character string regions, and detecting a white region whose pitch is different from the immediately precedent one as sub-separator.
Claim 5
Dependent Claim
70 words
The method according to claim 1, 2, 3 or 4 wherein if multiple patterns of object labels that may be assigned to said character strings are generated for a single model, said step (e) calculates a cost for each of said patterns on the basis of spatial relationships of said objects and said sub-separators and determines the most likely assignment pattern on the basis of the result of said calculation.
Claim 6
Dependent Claim
30 words
The method according to claim 5 wherein said cost calculation step imposes a penalty if character string regions to which the same object label is assigned continue beyond a sub-separator.
Claim 7
Dependent Claim
26 words
The method according to claim 2 wherein a sub-separator is treated as a separator unless a requisite number of separators are detected from said document image.
Claim 8
Independent Claim
59 words
The method of any one of the preceding claims wherein the registered layout model contains data on the spatial relationship of document objects in the model, data on whether each object is mandatory or not, data on the number of character strings which constitute a terminal node and data on the number of immediate successors for each intermediate node.
Claim 9
Independent Claim
110 words
A system for the layout analysis of an input document image comprised of objects arranged in a hierarchical tree structure, the system comprising: (a) means to register a layout model representing spatial relations of objects in a model document; (b) means to extract character string regions from the input document image; (c) means for detecting separators from said document image and generating data on their positions, said separators being white and/or black regions satisfying predetermined conditions; (d) means for segmenting said document image into tree-structured sub-areas by using said separator data and (e) means for matching said character string regions with said layout model by using said separators as boundaries.
Claim 10
Dependent Claim
48 words
The system of claim 9 further comprising means for detecting sub-separators within each of said sub-areas and generating data on their positions, said sub-separators being white regions satisfying predetermined conditions different from said separators, the matching means being adapted to employ said sub-separators as boundary candidates for objects.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.