Back to results
US 6,574,375 B1Granted
Method for detecting inverted text images on a digital scanning device
AssigneeRicoh Company, Ltd.
InventorsJohn F. Cullen, David G. Stork, Peter Hart, Koichi Ejiri
Life of this patent
Estimating the term…
Abstract
A method and apparatus for detecting proper page orientation of a scanned document image. The document is subdivided into a series of word boxes. The number of ascending and descending text characters within the word can then be compared with the number expected for a properly oriented document to verify page orientation.
Claims18
A method for determining whether a document image is inverted, the method comprising the steps of:
- identifying individual blocks of text, each block including at least one line of text within the document image, each of the individual blocks of text including a plurality of scan lines of pixel data;
- within at least one of the individual blocks of text, summing pixel data indicative of at least one feature in the text having a first of two possible logical values for each one of the plurality of scan lines to thereby obtain a sample data set including pixel distribution information for the at least one of the individual blocks of text;
- calculating a moment of the sample data set;
- determining if the document image is inverted by comparing the moment of the sample data set to at least a first stored normalized moment data set representing an image having a known upright orientation;
- and
- asserting an inversion signal if the step of comparing indicates the document image is inverted.
3 more elements
In a character recognition system, a method for determining inversion of a given image comprising the steps of:
- providing a representation of the image including one or more scan lines to a run length extraction and classification means;
- extracting run lengths from each scan line of the representation of the image;
- classifying each of the run lengths as short, medium or long based on a length of the run length, wherein a plurality of run length records are created;
- constructing rectangles from the run length information, the rectangles representing a portion of the image;
- classifying each of the rectangles as type image, vertical line, horizontal line, or unknown, the unknown rectangles primarily including individual text blocks;
- summing pixel data having a first of two possible logical values for each one of the plurality of scan lines within a given set of the individual text blocks to obtain a sample data set including information regarding pixel distribution within the given set of the individual text blocks;
- calculating a moment of the sample data set;
- comparing the moment of the sample data set to at least a first stored normalized moment data set representing an image having a known orientation to determine if the given image is inverted;
- automatically asserting an inversion signal if the step of comparing indicates the given image is inverted;
- and
- reorienting the given image in a non-inverted position when the inversion signal is asserted.
8 more elements
An apparatus for determining the orientation of a document image being processed, the apparatus comprising:
- means for identifying individual blocks of text within the document image, each of the individual blocks of text including a plurality of scan lines of pixel data;
- means for summing pixel data having a first of two possible logical values for each one of the plurality of scan lines within a given set of the individual blocks of text to obtain a sample data set including information regarding pixel distribution;
- means for calculating a moment of the sample data set;
- means for determining if the given document image is inverted by comparing the moment of the sample data set to at least a first stored normalized moment data set representing an image having a known orientation;
- and
- means for electronically reorienting the document image to an upright orientation when the means for comparing. determines the document image is inverted.
3 more elements
A photocopier comprising:
- means for automatically scanning a hard copy document to obtain a digital document image;
- a memory for storing at least a first set of normalized data for documents in a known orientation;
- and
- a processor, coupled to the means for scanning and to the memory, for:
- identifying individual blocks of text each including at least one line of text within the document image, each of the individual blocks of text including a plurality of scan lines;
- summing pixel data having a first of two possible logical values along the plurality of scan lines for at least one of the lines of text with the set of the individual blocks of text to obtain a sample data set, the data set including information regarding pixel distribution in the set of the individual blocks of text;
- calculating a moment of the sample data set;
- comparing the moment of the sample data set to the at least first set of normalized data retrieved from the memory to determine when the hard copy document is inverted;
- and
- automatically reorienting the document image to form a document image of upright orientation if the comparing function determines the hard copy document is inverted;
- and
- means for reorienting the digital document image to producing a hard copy representation of the hard copy document in its upright orientation.
9 more elements
An apparatus for automatically processing documents comprising:
- means for obtaining a digital document image;
- a memory for storing at least a first set of normalized data for documents in a known orientation;
- and
- a processor, coupled to the means for obtaining and to the memory, for:
- identifying individual blocks of text within the document image, each of the individual blocks of text including at least one line of text having a plurality of scan lines;
- summing pixel data having a first of two possible logical values for each one of the plurality of scan lines within at least one of the individual blocks of text to obtain a sample data set, the sample data set including pixel distribution information;
- calculating a moment of the sample data set;
- determining if the hard copy document is inverted by comparing the moment of the sample data set to the at least first set of normalized data retrieved from the memory;
- and
- electronically reorienting the document image to form an upright orientation document image when the hard copy document is determined to be inverted;
- and
- means for producing the upright orientation document image.
9 more elements
Dependencies are inferred from the claim text.