Word identifying device and method, and memory medium

A word recognition and word recognition technology, applied in the field of word recognition

Inactive Publication Date: 2002-05-01
FUJITSU LTD
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0011] However, since it is necessary to store about 4,000 kinds of character features of all characters whose positions and widths vary, a capacity of several hundred megabytes is required, which is a serious problem from a practical point of view

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Word identifying device and method, and memory medium
  • Word identifying device and method, and memory medium
  • Word identifying device and method, and memory medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0051] In the final stage of the feature extraction process described above, the weighted directionally coded histogram feature has a position at, for example, 7(long) for normalized character image segmentation * 8-directional features within a 7 (wide) grid, that is, weighted direction-encoded histogram features with 7 * 7 * 8-dimensional features. Here, 8 directions represent directions in units of 45° obtained by dividing 8 by 360°, as shown in FIGS. 2 and 3 .

[0052] In this preferred embodiment, the feature vectors are grouped in units of columns to reduce the capacity of the character feature dictionary.

[0053] Figure 5 A performance configuration diagram of the first preferred embodiment is shown.

[0054] In the figure, when learning, the feature vector extracted from the input character image is stored in the character feature dictionary.

[0055] The grouping unit 12 related to this preferred embodiment is the character feature stored in the character featu...

Embodiment 2

[0096] Next, a description will be given of a second embodiment according to the present invention, wherein after the column features are grouped, the capacity of the feature dictionary is reduced using a combination coefficient.

[0097] Suppose the number of encoded column vectors (representative vectors) is m and the "p"th column vector is f p , and the composite coefficient is k i . At this time, it is checked whether there is a combination of the combination coefficient k and the column vector, which can be represented by the following formula (1). If there is a corresponding combination, the identification number of the column vector and the combination coefficient are registered. f p = Σ i m k i * f i ( i ≠ p ) ...

Embodiment 3

[0114] For a weighted directionally encoded histogram feature, in order to reduce the redundancy of information contained in the feature, by extracting 7 * 7 * The 8-dimensional initial features are subjected to feature transformation, such as standard discriminant analysis, etc., to achieve dimensionality compression. As a result, the feature dimensionality drops, for example, from 392 to about 100. As mentioned above, features previously transformed with feature transformations such as principal element analysis, standard discriminant analysis, etc. are grouped and encoded, thereby reducing the size of the dictionary.

[0115] Figure 15 A performance configuration diagram of the third preferred embodiment is shown.

[0116] exist Figure 15 As described above, feature vectors extracted from input character images are stored in the feature dictionary 11 at the time of learning.

[0117] The capacity reduction unit 31 related to the present preferred embodiment is config...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The capacity of a character feature dictionary is reduced, and stored as a feature dictionary. The capacity is reduced by clustering feature vectors in units of columns or rows for character features, by making m column vectors represent the column or row features, and by assigning 1 to m identification numbers. The capacity of the dictionary can be further reduced by representing a column or row feature with an addition sum of other column or row features, or differential features after clustering is performed, or by performing dimension compression for character features. Word recognition is performed by synthesizing a word feature for a comparison based on a word list to be recognized, and by making a comparison between a feature extracted from an input word and the synthesized feature. Or, a comparison between input word and input word features whose numbers of dimensions are different may be made with nonlinear elastic matching.

Description

technical field [0001] The present invention relates to character recognition in character recognition devices. Background technique [0002] In recent years, there has been an increasing demand for character recognition devices OCR (Optical Character Recognition) or software OCR. [0003] Character recognition is a method by which handwritten characters such as "Tokyo" are not separated into individual characters for recognition, but the character itself is recognized as a whole. In this way, high-precision recognition can be achieved even when there is contact between characters. This is one of the effective methods for recognizing handwritten character strings in the free character spacing area. A character recognition device according to the present invention can be applied not only to handwritten character recognition devices, but also to generalized character recognition devices, such as printed character recognition devices, character recognition devices for portabl...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06V30/10
CPCY10S707/99936Y10S707/99932G06K9/723G06V30/268G06V30/10
Inventor 堀田悦伸
Owner FUJITSU LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap