This class contains data associated with characters in OCRMax. This includes the character's character code, the character's image, rectangles that describe where the character appears in the image, and other data.

The CogOCRMaxChar class holds an object of type CogOCRMaxCharKey in its Key property. The Key contains information that identifies the CogOCRMaxChar. Most importantly the Key property holds the CharacterCode. The CharacterCode is a Unicode UTF-32 value which determines the 'letter' a CogOCRMaxChar corresponds to.

A collection of CogOCRMaxCharKey objects.

This class contains metadata associated with characters for OCRMax.

Classification is the second main step in performing OCRMax. Classification takes in a CogOCRMaxChar produced by the segmenter and returns the best matching character, the score of that best matching character, and the confidence of that best matching character. By classifying all of the segmented characters in a line of text, the classifier determines the result string for the whole line. See the OCRMax Theory Guide for more information.

Result data for one potential classification of the run-time image.

A Collection of CogOCRMaxClassifierCharResults.

All result data for the classification of one line of characters.

All result data for the classification of a run-time character.

Run-time parameters for the OCRMax Classifier.

Training parameters for the OCRMax Classifier.

Fielding is the third main step in performing OCRMax.

Fielding takes the Classifier results as an input and outputs a refined set of results based on prior information and expectations of the text being read. For instance, if you know the text being read always start with the letter 'C', fielding can use this information to improve the classifier results.

The Classifier results represent a line of character positions and the scores of different characters at each position.

A user configures the Fielding to provides information about what characters are expected at different positions in the string. Fielding uses this information, and the information in the Classifier results, to make small corrections in situations where the Classifier was unable to make a clear determination of what character was printed at a particular position. Fielding may also use the expected string information to invalidate the classifier result at locations where the classifier found something very different than what was expected by the fielding.

This class is a named, reusable description of the characters that are expected at some position. Multiple character positions may be specified by referring to the name, or 'alias', of a single field definition object.

This class holds zero or more CogOCRMaxFieldingDefinition objects. Note that the ordering of objects within this collection does not affect processing.

The result of fielding a line of text.

A verification result for a single position.

A container of CogOCRMaxChars (character images and their associated character codes) used to train the OCRMax classifier.

A Collection of CogOCRMaxPositionResults which represent the output/result of running the OCRMax tool.

Contains a the result for a single character position in a CogOCRMaxLineResult.

CogOCRMaxSegmenter performs the first main step of OCRMax. Segmentation takes an image containing a line of text as input, and outputs CogOCRMaxChar(s) structures which contain information about where individual characters can be found in the image. For example, if the input image contains the line of text "ABCD", then segmenter outputs four separate CogOCRMaxChars, each of which contains an image of a single character. The segmenter does not know anything about actually determining which letter a CogOCRMaxChar corresponds to; in the above example, it has no way of knowing that the CogOCRMaxChar is the letter "A". The job of identifying each CogOCRMaxChar is left to the classifier. See the OCRMax Theory Guide for more information.

Statistical information about a set of OCRMax Segmenter characters

Information about an OCR element, which could be a fragment or character. One or more fragments joined together constitute a character. A fragment can also get split into more than one character.

A collection of CogOCRMaxSegmenterElementInfos

Statistical information about a set of OCRMax Segmenter fragments

A segmenter line result represents an entire line of segmented characters.

Contains information about how the OCRMax Segmenter arrived at a particular line result.

A segmentation paragraph result contains one or more lines of segmented characters.

A segmentation position result represents a single segmented character.

A segmenter result is a collection of one or more CogOCRMaxSegmenterParagraphResults. This result is returned when using the Execute override that accepts multiple regions of interest as inputs.

Parameters to control how spaces between characters are handled by the segmenter.

This class defines a single group of swap characters. Characters within the group are swappable with each other. (They appear similar and are hard to distinguish, e.g. '0' (zero) and 'O' (capital oh)).

This class is a container that holds multiple CogOCRMaxSwapChars. Each CogOCRMaxSwapChar holds a set of character codes that are "swappable" with each other.

The CogOCRMaxTool (Optical Character Recognition) reads a line of text in an image. The CogOCRMaxTool performs three primary operations: segmentation, classification, and fielding. Segmentation occurs first and uses thresholding techniques to identify which areas in the image appear to contain lines of text. Classification occurs second and is responsible for "reading" any text found by the segmenter. It does this by comparing the images of segmented characters to previously acquired images of known characters in a font. Fielding is the third and final stage of OCRMax. Fielding is an optional step which uses prior information about the text being read to improve or correct the classification results. For more details see the documentation for CogOCRMaxSegmenter, CogOCRMaxClassifier, CogOCRMaxFielding, as well as the OCRMax Theory Guide.

CogOCRMaxTuneData contains a set of CogOCRMaxTuneRecords for use with training the CogOCRMaxTool.

Tune parameters for the CogOCRMaxTool.

Contains data about a line of text in an image that can be used to tune the CogOCRMaxTool.

A collection of CogOCRMaxTuneRecords

A collection of CogRectangleAffine objects.

Cognex Custom Character Codes

The algorithm to use for classification

The Aspect Mode to use for classification

The reason a classified character was confused

What image preprocessing to perform before classification. Note that if both are enabled, histogram normalization happens after median subtraction.

The status of a classified character

Enumeration of data that can be provided in the Tool's CurrentRecord.

How the CogOCRMaxFielding operator will judge Accepted.

How confidence/confusion will be computed by the CogOCRMaxFielding operator.

Status values returned by fielding to indicate the state of a fielded position result.

Enumeration of non-diagnostic data that can be provided in the Tool's LastRunRecord.

Enumeration of diagnostic data that can be provided in the Tool's LastRunRecord.

Conveys information about the state of an OCRMax Tool's Line Result. Returns the most severe status found in the line result's position results.

Enumeration of character polarity constants.

Conveys information about the state of an Position Result

Enumeration of data used for creating result graphics.

Type of analysis to perform in order to determine the optimal character segmentation.

Used to describe how OCRMax will compute the width of the cell rectangle that surrounds each character.

How to merge character fragments to form characters during the Group stage. Note that the Analyze stage may perform additional merges.

Provides diagnostic information about segmenter characters and fragments in order to show how the segmenter arrived its results.

Used to describe the width of characters for OCRMax.

Describes the pitch of characters for OCRMax.

Used to describe the pitch of characters for OCRMax.

Describes the method of line refinement for OCRMax character segmentation.

How to normalize the image so that the normalized image can be binarized well using only a global threshold.

How to handle insertion of space characters into gaps between other characters.

How to compute the score of a space character.

Controls which CogOCRMaxSegmeter properties will be tunned.

Identifies tune record for use with OCRMax tuning was created.