Assembly: Cognex.VisionPro.OCRMax (in Cognex.VisionPro.OCRMax.dll) Version: 65.1.0.0
Parameters
- image
- Type: Cognex.VisionPro CogImage16Range
The Image from which to segment characters.
- region
- Type: Cognex.VisionPro CogRectangleAffine
The region of interest (ROI) of the image from which to segment characters.
Return Value
Type: CogOCRMaxSegmenterParagraphResult| Exception | Condition |
|---|---|
| CogSecurityViolationException | |
| CogOCRMaxSegmenterParamsException |
Thrown if:
|
| CogOCRMaxSegmenterParamsException |
Thrown if:
|
| CogTransformNotLinearException |
Thrown if:
|
| CogImageAllocatedException | Thrown if the image is not allocated. |
| CogTimeoutExpiredException | Thrown if the Timeout expired. |
| ArgumentNullException | Thrown if image is null. |
| NotImplementedException | Thrown if LineRefinementMode is set to Classic. |
Segmentation is performed on the 16 bit grey pixel data after the removal of missing pixels. If the input range image GetMaskData() returns a null visible pixel mask it is assumed that any missing pixels were removed external to the segmenter in a previous step. If the image's visible pixel mask is _not_ null, the segmenter removes the missing pixels using a neighborhood interpolation filter of the image within the region before performing segmentation.
The region should contain one line of characters and surrounding featureless, but possibly nonuniform and/or noisy, background; there should be no strong features in the region other than the one line of characters.
The region is specified in the selected space coordinates of the input image. The x-axis should be approximately parallel to the baseline of the contained line of characters, with its positive direction pointing along the reading direction. The skew should be the approximate value of the skew of the line of characters.
If AngleHalfRange is greater than zero, the angle of the line will be refined. If SkewHalfRange is greater than zero, the skew of the line will be refined.