CHARACTER SEGMENTATION DEVICE, CHARACTER RECOGNITION DEVICE AND CHARACTER SEGMENTATION METHOD

• US 20160292528A1
• Filed: 03/30/2016
• Published: 10/06/2016
• Est. Priority Date: 03/31/2015
• Status: Active Grant

• Current Assignee: Nidec Sankyo Corporation (Nidec Corporation)
• Original Assignee: Nidec Sankyo Corporation (Nidec Corporation)
• Inventors: NAKAMURA, Hiroshi
• Granted Patent: US 9,886,643 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/162 Quantising the image signal

Abstract

A character recognition device may include an image sensor to acquire image and a data processor configured to segment a character string from the image data. The data processor may be configured to detect positions to segment the characters; set an area to detect a position for segmenting characters; create a first projection of pixel values arranged in the character arrangement direction along which characters are arranged and a second projection of pixel values arranged in an orthogonal to said character arrangement direction; calculate a synthetic threshold value of a first threshold, acquired by applying a discriminant analysis method on said first, X-axis projection, and a second threshold, acquired by applying said discriminant analysis method on said second, Y-axis projection, and take it as a binary threshold for said setting area; and acquire character segmenting positions based on said binary threshold.

Claims

1. A character recognition device for use with a recording medium comprising a character string comprising characters, the character recognition device comprising:

an image sensor configured to image the character string on the recording medium and read the character string as image data; and

a data processor configured to segment a character string from said image data, and segment the characters from said character string to implement character recognition;

wherein said data processor is further configured to detect positions to segment the characters;

said data processor is further configured to;

set an area to detect a position for segmenting said characters;

create a first projection of pixel values of the pixels arranged in the character arrangement direction along which characters are arranged and a second projection of pixel values of the pixels arranged in an orthogonal to said character arrangement direction, at least within said setting area;

calculate a synthetic threshold value of a first threshold, acquired by applying a discriminant analysis method on said first, X-axis projection, and a second threshold, acquired by applying said discriminant analysis method on said second, Y-axis projection, and take it as a binary threshold for said setting area; and

acquire character segmenting positions based on said binary threshold.

2. A character segmentation method for use with image data acquired by capturing a character string comprising characters on a recording medium, the character segmentation method comprising:

imaging the character string on the recording medium with an image sensor;

a character segmenting position-detecting step comprising detecting segmenting-positions of the characters in the character string imaged by the image sensor;

wherein said character segmenting position-detecting step further comprises;

an area-setting step comprising setting an area to detect positions for segmenting said character;

a projection-creating step comprising creating a first projection of pixel values of the pixels arranged in the character arrangement direction along which characters are arranged and a second projection of pixel values of the pixels arranged in an orthogonal to said character arrangement direction, at least within a setting area set by said area-setting section;

a binary threshold-acquiring step comprising calculating a synthetic threshold value of a first threshold, acquired by applying a discriminant analysis method on said first, X-axis projection and a second threshold, acquired by applying said discriminant analysis method on said second, Y-axis projection and takes it as a binary threshold for said setting area; and

a position-detecting step comprising acquiring character segmenting positions based on said binary threshold acquired at said binary threshold-acquiring section.

3. A non-transitory computer-readable medium for use with an image sensor and a recording medium comprising a character string comprising characters, the non-transitory computer-readable medium storing computer-executable instructions that, when executed by a computer, cause the computer to perform:

imaging the character string on the recording medium with an image sensor;

a character segmenting position-detecting step comprising detecting segmenting-positions of the characters in the character string imaged by the image sensor;

wherein said character segmenting position-detecting step further comprises;

an area-setting step comprising setting an area to detect positions for segmenting said character;

a projection-creating step comprising creating a first projection of pixel values of the pixels arranged in the character arrangement direction along which characters are arranged and a second projection of pixel values of the pixels arranged in an orthogonal to said character arrangement direction, at least within a setting area set by said area-setting section;

a binary threshold-acquiring step comprising calculating a synthetic threshold value of a first threshold, acquired by applying a discriminant analysis method on said first, X-axis projection and a second threshold, acquired by applying said discriminant analysis method on said second, Y-axis projection and takes it as a binary threshold for said setting area; and

a position-detecting step comprising acquiring character segmenting positions based on said binary threshold acquired at said binary threshold-acquiring section.