Class OcrInputBase

Base class for OCR input variants

Inheritance

System.Object

OcrInputBase

OcrImageInput

OcrInput

OcrPdfInput

Implements

System.IDisposable

Namespace: IronOcr

Assembly: IronOcr.dll

Syntax

public abstract class OcrInputBase : Object, IDisposable

Constructors

OcrInputBase()

Declaration

protected OcrInputBase()

Properties

TargetDPI

The resolution that low resolution images will be enhanced to. To disable upscaling, set this to 0 (will affect read quality).

TargetDPI also determines the resolution at which PDF documents will be sampled.

Declaration

public int TargetDPI { get; set; }

Property Value

Type	Description
System.Int32

Title

A title for the OcrInput document. This is relevant as it becomes metadata when exporting searchable PDFs and HOCR files from IronTesseract results.

See SaveAsSearchablePdf(String) and SaveAsHocrFile(String)

Declaration

public string Title { get; set; }

Property Value

Type	Description
System.String

Methods

AdaptiveThreshold(Nullable<Single>)

Applies Bradley Adaptive Threshold to the image.

Adaptive thresholding is the method where the threshold value is calculated for smaller regions and therefore, there will be different threshold values for different regions.

Declaration

public OcrInputBase AdaptiveThreshold(Nullable<float> thresholdLimit = null)

Parameters

Type	Name	Description
System.Nullable<System.Single>	thresholdLimit	Threshold limit (0.0-1.0) to consider for binarization. 0.0 to consider threshold is completely white 1.0 to consider threshold is completely black

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

ApplyMultipleFilters(OcrFilters, Double, Int32, Int32, Int32, Boolean, Nullable<Int32>)

Apply multiple imaging filters using the specified parameters

Declaration

public void ApplyMultipleFilters(OcrFilters filters, double Rotation = 0, int MaxDeskewAngle = 45, int MaxWidth = 0, int MaxHeight = 0, bool Use3x3 = false, Nullable<int> thresholdLimit = null)

Parameters

Type	Name	Description
OcrFilters	filters	Filters to apply
System.Double	Rotation	Rotation amount. Required when using the Rotation filter. Rotate
System.Int32	MaxDeskewAngle	Optional MaxDeskewAngle amount when using the Deskew filter. Defaults to 45 degrees.Deskew
System.Int32	MaxWidth	Maximum width. Required when using the Scale filter. Scale
System.Int32	MaxHeight	Maximum height. Required when using the Scale filter. Scale
System.Boolean	Use3x3	Optional morphology when using the Despeckle (DeNoise), Dilate, or Erode filter. DeNoise, Dilate, or Erode
System.Nullable<System.Int32>	thresholdLimit	Optional Threshold limit (0.0-1.0) to consider for binarization when using the Bradley Adaptive Threshold.

Remarks

This method serves as an alternative way to apply multiple filters. Filters are applied in what is typically the optimal order.

Exceptions

Type	Condition
System.ArgumentOutOfRangeException

Binarize()

This image filter turns every pixel black or white with no middle ground. May Improve OCR performance cases of very low contrast of text to background.

Declaration

public OcrInputBase Binarize()

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Close(Boolean)

Advanced Morphology.

Closing is reverse of Opening, Dilation followed by Erosion. It is useful in closing small holes inside the foreground objects.

Declaration

public OcrInputBase Close(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Contrast(Single)

Increases contrast automatically. This filter often improves OCR speed and accuracy in low contrast scans. Flattens Alpha channels to white.

Declaration

public OcrInputBase Contrast(float amount = 1.1F)

Parameters

Type	Name	Description
System.Single	amount	Amount which is used to adjust contrast. A value of 0 will create an image that is completely gray. A value of 1 leaves the input unchanged. Amount values greater than 0 increase contrast making light areas lighter and dark areas darker. Amount values less than 0 decrease contrast - decreasing variety of contrast.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

DeNoise(Boolean)

Removes digital noise. This filter should only be used where noise is expected. Flattens Alpha channels to white.

Declaration

public OcrInputBase DeNoise(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

See Also

Despeckle(Boolean)

Deskew(Int32)

Rotates an image so it is the right way up and orthogonal. This is very useful for OCR because Tesseract tolerance for skewed scans can be as low as 5 degrees.

This also helps when producing searchable PDF documents from IronTesseract because the pages will likely all be the right way up.

This version uses only Hough transform to make minor correction. Example: pages that where put in a scanner at a slight angle.

Declaration

public bool Deskew(int MaxDeskewAngle = 45)

Parameters

Type	Name	Description
System.Int32	MaxDeskewAngle	Maximum angle of skew to correct for. Higher values can lead to more opportunity for correction, but may be slower and more prone to error including upside down pages.

Returns

Type	Description
System.Boolean	Returns a boolean result of whether or not IronOCR was able to detect image orientation. True = Deskew was applied. False = Failed to detect image orientation and image remains unchanged.

See Also

Rotation

Despeckle(Boolean)

DeSpeckle as an alias of DeNoise.

Alias of DeNoise(Boolean) to make this method easily to find in Intensense.

Declaration

public OcrInputBase Despeckle(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

DetectPageOrientation()

Detects the page orientation for each image in a collection using specified detection modes.

Declaration

public IEnumerable<OcrPageOrientationResult> DetectPageOrientation()

Returns

Type	Description
System.Collections.Generic.IEnumerable<OcrPageOrientationResult>	An array of OcrPageOrientationResult where each element represents the detected orientation for an image in the collection.

Remarks

Higher text density within the image or PDF generally leads to more accurate orientation detection.

The correctional rotation angle can be directly used to rotate the image to the correct orientation.

Dilate(Boolean)

Advanced Morphology. Dilation is the opposite of Erosion, instead of shrinking it expands the foreground object.

Opposite of Erode(Boolean).

Declaration

public OcrInputBase Dilate(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Dispose()

Disposes of this OcrInput object.

Declaration

public void Dispose()

EnhanceResolution(Int32)

Enhances the resolution of low quality images. This filter is not often needed because TargetDPI will automatically catch and resolve low resolution inputs.

May not work for all images if their metadata is corrupted.

Declaration

public OcrInputBase EnhanceResolution(int TargetDPI = 225)

Parameters

Type	Name	Description
System.Int32	TargetDPI	The target DPI to resample to.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Erode(Boolean)

Advanced Morphology. Erosion is the morphological operation used to diminish the size of the foreground object.

Opposite of Erode(Boolean).

Declaration

public OcrInputBase Erode(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Finalize()

OcrInput has a safe finaliser that cleans up undisposed native images in memory.

Declaration

protected override void Finalize()

FindMultipleTextRegions(Double, Int32, Boolean, Boolean)

Use computer vision to detect areas which contain text elements and divide the input into separate images based on text regions.

Declaration

public void FindMultipleTextRegions(double Scale = 0, int DilationAmount = -1, bool Binarize = true, bool Invert = false)

Parameters

Type	Name	Description
System.Double	Scale	(Only used during text region detection) Resolution scale factor. Image width and height will be multiplied by this value.
System.Int32	DilationAmount	(Only used during text region detection) Dilation amount, in pixels. Text areas width and height will be increased by this value.
System.Boolean	Binarize	(Only used during text region detection) True to convert the image to black and white, False otherwise
System.Boolean	Invert	(Only used during text region detection) True to invert image colors during binarization, False otherwise

Remarks

Useful for generating several OCR results from a single image/page

FindTextRegion(Double, Int32, Boolean, Boolean)

Use computer vision to detect regions which contain text elements on each page

Declaration

public void FindTextRegion(double Scale = 0, int DilationAmount = -1, bool Binarize = true, bool Invert = false)

Parameters

Type	Name	Description
System.Double	Scale	(Only used during text region detection) Resolution scale factor. Image width and height will be multiplied by this value.
System.Int32	DilationAmount	(Only used during text region detection) Dilation amount, in pixels. Text areas width and height will be increased by this value.
System.Boolean	Binarize	(Only used during text region detection) True to convert the image to black and white, False otherwise
System.Boolean	Invert	(Only used during text region detection) True to invert image colors when binarizing, False otherwise

GetPages()

Get all OcrInputPage within this OcrInput

Declaration

public IEnumerable<OcrInputPage> GetPages()

Returns

Type	Description
System.Collections.Generic.IEnumerable<OcrInputPage>

HighlightTextAndSaveAsImages(IronTesseract, String, ResultHighlightType)

Based on the ResultHighlightType, will draw red boxes around characters/words/lines/paragraphs detected, and save to a PNG image.

For best results, perform all filters before calling.

Declaration

public void HighlightTextAndSaveAsImages(IronTesseract tesseract, string filename, ResultHighlightType type)

Parameters

Type	Name	Description
IronTesseract	tesseract	IronTesseract instance used to scan the OcrInput.
System.String	filename	File will be saved as : 'filename_page_0.png'. You may use an absolute or relative path.
ResultHighlightType	type	Choose whether each box represents a character, word, line, paragraph.

HoughTransformStraighten(Int32)

Uses a Hough Transform to rotate and image to the nearest 90 degrees of straightness. This is very useful for OCR because Tesseract tolerance for skewed scans can be as low as 5 degrees.

A Synonym of Deskew(Int32)

Declaration

public OcrInputBase HoughTransformStraighten(int MaxDeskewAngle = 45)

Parameters

Type	Name	Description
System.Int32	MaxDeskewAngle	Maximum angle of skew to correct for. Higher values can lead to more opportunity for correction, but may be slower and more prone to error including upside down pages.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

See Also

Rotation

Invert(Boolean)

Inverts every color. E.g. White becomes black : black becomes white.

Declaration

public OcrInputBase Invert(bool GrayScale = false)

Parameters

Type	Name	Description
System.Boolean	GrayScale	Optionally remove all color channels and return a GrayScale image.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Open(Boolean)

Advanced Morphology.

Opening is just another name of erosion followed by dilation. It is useful in removing noise.

Declaration

public OcrInputBase Open(bool use3x3 = false)

Parameters

Type	Name	Description
System.Boolean	use3x3	2x2 is default morphology

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

PageCount()

The number of OcrInputPages currently present in this OcrInput

Declaration

public int PageCount()

Returns

Type	Description
System.Int32

RemovePage(Int32)

Removes a OcrInputPage within this OcrInput at the specified index.

Declaration

public bool RemovePage(int pageIndex)

Parameters

Type	Name	Description
System.Int32	pageIndex	Indexes of a page in this OcrInput to remove.

Returns

Type	Description
System.Boolean	True if page was removed sucessfully. False if removal failed.

RemovePages(IEnumerable<Int32>)

Removes all OcrInputPage within this OcrInput at the specified indexes.

Declaration

public bool RemovePages(IEnumerable<int> pageIndexes)

Parameters

Type	Name	Description
System.Collections.Generic.IEnumerable<System.Int32>	pageIndexes	Indexes of pages in this OcrInput to remove.

Returns

Type	Description
System.Boolean	True if ALL page indexes were removed sucessfully. False if one or more failed.

ReplaceColor(Color, Color, Int32)

Replace current color to new color in Image

Declaration

public OcrInputBase ReplaceColor(Color currentColor, Color newColor, int tolerance = 10)

Parameters

Type	Name	Description
IronSoftware.Drawing.Color	currentColor	Current IronSoftware.Drawing.Color
IronSoftware.Drawing.Color	newColor	New IronSoftware.Drawing.Color
System.Int32	tolerance	Tolerance Value

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Rotate(Double)

Rotates images by a number of degrees clockwise. For anti-clockwise, use negative numbers. Also see Deskew(Int32)

Declaration

public OcrInputBase Rotate(double Degrees)

Parameters

Type	Name	Description
System.Double	Degrees	A number of clockwise degrees.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

SaveAsImages(String, AnyBitmap.ImageFormat)

Exports an OcrInput object as Images

Declaration

public string[] SaveAsImages(string Prefix = "export_of_page", AnyBitmap.ImageFormat Extension)

Parameters

Type	Name	Description
System.String	Prefix	Will save images at {Prefix}_(page_number).{Extension}. May include a fully qualified file path.
IronSoftware.Drawing.AnyBitmap.ImageFormat	Extension	Output file extension in lower-case.

Returns

Type	Description
System.String[]	Array of saved image file names. Can be multiple if OcrInput used has multiple pages.

Exceptions

Type	Condition
System.Exception	Throws an exception if there are no pages. See OcrInputPage

Scale(Int32, Boolean)

Scales OCRInput pages proportionally.

Declaration

public OcrInputBase Scale(int Percentage, bool ScaleCropArea = true)

Parameters

Type	Name	Description
System.Int32	Percentage	The percentage scale. 100 = no effect.
System.Boolean	ScaleCropArea	Should associated crop areas also be scaled proportionally (recommended true)

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Scale(Int32, Int32, Boolean)

Scales the OCRInput pages up in size.

Declaration

public OcrInputBase Scale(int MaxWidth, int MaxHeight, bool ScaleCropArea = true)

Parameters

Type	Name	Description
System.Int32	MaxWidth	Maximum width in pixels.
System.Int32	MaxHeight	Maximum height in pixels.
System.Boolean	ScaleCropArea	Should associated crop areas also be scaled proportionally (recommended true)

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

SelectTextColor(Color, Int32)

Binarize an image to read pixels of a color (with threshold) as text and ignore other colors as background.

This is useful if you image has many colors and a normal binarize will not work. It will turn all text of the color specified into black and the rest of the image to white.

Declaration

public OcrInputBase SelectTextColor(Color selectColor, int tolerance = 10)

Parameters

Type	Name	Description
IronSoftware.Drawing.Color	selectColor	IronSoftware.Drawing.Color of text to isolate from background.
System.Int32	tolerance	[0,255]; Acceptable range of the difference between PixelColor and selectColor for each R, G, and B value

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

SelectTextColors(IEnumerable<Color>, Int32)

Binarize an image to read pixels of only selected colors (with thresholds) as text and ignore other colors as background.

This is useful if you image has many colors and a normal binarize will not work. It will turn all text of the color specified into black and the rest of the image to white.

Declaration

public OcrInputBase SelectTextColors(IEnumerable<Color> selectColors, int tolerance = 10)

Parameters

Type	Name	Description
System.Collections.Generic.IEnumerable<IronSoftware.Drawing.Color>	selectColors	IronSoftware.Drawing.Color of text to isolate from background.
System.Int32	tolerance	[0,255]; Acceptable range of the difference between PixelColor and selectColor for each R, G, and B value

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Sharpen()

Sharpens blurred OCR Documents. Applies a Gaussian sharpening filter to image.

Declaration

public OcrInputBase Sharpen()

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

StampCropRectangle(OcrInput, Rectangle, Color)

Returns a new OcrInput object with a rectangle applied and to visualize and debug where a Crop Rectangle will be applied when using IronSoftware.Drawing.Rectangle on your image.

Declaration

public static OcrInput StampCropRectangle(OcrInput input, Rectangle cropRectangle, Color rectangleColor)

Parameters

Type	Name	Description
OcrInput	input	OcrInput object to return a copy of stamped
IronSoftware.Drawing.Rectangle	cropRectangle	Use a IronSoftware.Drawing.Rectangle to debug the area that will be scanned on an image.
IronSoftware.Drawing.Color	rectangleColor	Color of rectangle drawn. Red is recommended for easy contrast.

Returns

Type	Description
OcrInput

StampCropRectangleAndSaveAs(Rectangle, Color, String, AnyBitmap.ImageFormat)

Saves a copy of the image with a rectangle applied and to visualize and debug where Crop Rectangle will be applied when using IronSoftware.Drawing.Rectangle on your image.

Declaration

public string[] StampCropRectangleAndSaveAs(Rectangle cropRectangle, Color rectangleColor, string Prefix = "rectangle_on_page", AnyBitmap.ImageFormat extension)

Parameters

Type	Name	Description
IronSoftware.Drawing.Rectangle	cropRectangle	Use a IronSoftware.Drawing.Rectangle to debug the area that will be scanned on an image.
IronSoftware.Drawing.Color	rectangleColor	Color of rectangle drawn. Red is recommended for easy contrast.
System.String	Prefix	Will save images at {Prefix}_(page_number).{Extension}. May include a fully qualified file path.
IronSoftware.Drawing.AnyBitmap.ImageFormat	extension	Output file extension in lower-case.

Returns

Type	Description
System.String[]	Array of saved image file names. Can be multiple if OcrInput used has multiple pages.

Exceptions

Type	Condition
System.Exception	Throws an exception if there are no pages. See OcrInputPage

StampCropRectangles(OcrInput, Rectangle[], Color)

Returns a new OcrInput object with a rectangle applied and to visualize and debug where the Rectangle will be applied when using IronSoftware.Drawing.Rectangle on your image.

Declaration

public static OcrInput StampCropRectangles(OcrInput input, Rectangle[] cropRectangles, Color rectangleColor)

Parameters

Type	Name	Description
OcrInput	input	OcrInput object to return a copy of stamped
IronSoftware.Drawing.Rectangle[]	cropRectangles	Use an array of IronSoftware.Drawing.Rectangle to debug the areas that will be scanned on an image.
IronSoftware.Drawing.Color	rectangleColor	Color of rectangle drawn. Red is recommended for easy contrast.

Returns

Type	Description
OcrInput

StampCropRectanglesAndSaveAs(Rectangle[], Color, String, AnyBitmap.ImageFormat)

Saves a copy of the image with all rectangles applied and to visualize and debug where Crop Rectangles will be applied when using IronSoftware.Drawing.Rectangle on your image.

Declaration

public string[] StampCropRectanglesAndSaveAs(Rectangle[] cropRectangles, Color rectangleColor, string Prefix = "rectangle_on_page", AnyBitmap.ImageFormat extension)

Parameters

Type	Name	Description
IronSoftware.Drawing.Rectangle[]	cropRectangles	Use an array of IronSoftware.Drawing.Rectangle to debug the areas that will be scanned on an image.
IronSoftware.Drawing.Color	rectangleColor	Color of rectangles drawn. Red is recommended for easy contrast.
System.String	Prefix	Will save images at {Prefix}_(page_number).{Extension}. May include a fully qualified file path.
IronSoftware.Drawing.AnyBitmap.ImageFormat	extension	Output file extension in lower-case.

Returns

Type	Description
System.String[]	Array of saved image file names. Can be multiple if OcrInput used has multiple pages.

Exceptions

Type	Condition
System.Exception	Throws an exception if there are no pages. See OcrInputPage

ToGrayScale()

This image filter turns every pixel into a shade of grayscale. Unlikely to improve OCR accuracy but may improve speed.

Declaration

public OcrInputBase ToGrayScale()

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

WithTitle(String)

Adds a Title to the OcrInput Document. This title will be used when calling SaveAsHocrFile(String) and SaveAsSearchablePdf(String)

Declaration

public OcrInputBase WithTitle(string Title)

Parameters

Type	Name	Description
System.String	Title	The document title as a string.

Returns

Type	Description
OcrInputBase	This OcrInput object allowing for LINQ style fluent notation.

Implements

System.IDisposable