interface IImageReader in Tinman.Terrain.Imaging

Base interface for classes that read images sequentially by scanline.

interface	IImageReader	extends	IDisposable
			IImageInfo
			IImageOps<IImageReader>
			IValidatable
		base of	IHeightmapReader
			ImageReaderBase

Remarks

When an image is read, the following processing is performed:

1. Images are read one scanline at a time (see Scanline), starting with the scanline at image pixel Y-coordinate zero. The color channels ( Red, Green, Blue, Intensity) and / or alpha channel (Alpha) of each raw pixel in a scanline are processed as described below.
2. Void regions (see NoData):
   If any channel value of the raw pixel is equal to the no-data value, the output pixel is set to void (see Set_Void) and further processing is skipped.
3. Range normalization (see Range):
   A linear transformation (see Transform) is applied to all channel values of the raw pixel, outputting a normalized pixel.
4. Gamma correction (see Gamma):
   The color channel values of the normalized pixel are fed into the gamma correction formula, producing the color channel values of the linear pixel. Gamma correction is not applied to alpha channel values.
5. The channel values of the linear pixel are clamped to the range [0..16777215] and round to 24-bit integer values (see Pixels).

Attributes

Coverage

The pixel coverage of the image.

property	Coverage { get }
type	PixelCoverage
value		The pixel coverage.
inherited	IImageInfo.Coverage

Remarks:

Usually, images have a pixel coverage of PixelIsArea.

Gamma

The gamma correction that must be applied to normalized image pixels in order to obtain linear pixel values.

property	Gamma { get }
type	float64
value		The gamma correction value (see remarks).
inherited	IImageInfo.Gamma

Remarks:

Depending on the value V of this property, the following gamma correction is applied:

• V > 0
  Given a normalized image pixel value N, the resulting linear pixel value L is computed as L = pow(N, V).
• V = -2.2 (see GammaFromSrgb)
  The normalized image pixel value is fed into the FromSrgbF method and its return value is used as the linear pixel value.
• V = -0.45455 (see GammaToSrgb)
  The normalized image pixel value is fed into the ToSrgbF method and its return value is used as the linear pixel value (which is not correct in a mathematical sense, but might be necessary in rare cases).
• For all other values of V, the normalized and linear pixel values are treated as being equivalent.

See also:

IImageReader

Georef

Returns georeferencing information.

property	Georef { get }
type	Raster
value		The georeferencing object or null if no georeferencing is available.
inherited	IGeorefInfo.Georef

HasFinished

Has the image reader finished (i.e. the whole image has been read)?

property	HasFinished { get }
type	bool
value		true if the whole image has been read, false if not.

Height

Height of the image, in pixels.

property	Height { get }
type	int32
value	[>0]	Height of the image, in pixels.
inherited	IImageInfo.Height

Remarks:

This is the number of image pixels along the Y-axis, which may be different from the raster height (see MapInfo), depending on the pixel coverage.

ImageFormat

The file format of the image file.

property	ImageFormat { get }
type	ImageFormat
value	[not-null]	The image file format.
inherited	IImageInfo.ImageFormat

Remarks:

The default file format is Any, which will try to load image data using any of the available image formats.

Layer

The heightmap layer which shall receive the data of this image.

property	Layer { get }
type	HeightmapLayer
value		The target heightmap layer.
inherited	IImageInfo.Layer

Remarks:

This is not an actual property of the image, but included here for convenience when images are imported into heightmap datasets.

LifecycleState

Returns the lifecycle state of this object.

property	LifecycleState { get }
type	LifecycleState
value		The lifecycle state.
inherited	ILifecycleState.LifecycleState

Map

Returns a MapInfo value that describes the used map raster.

property	Map { get }
type	MapInfo
value		The map info value.
inherited	IMapInfo.Map

NoData

The raw pixel value that represents missing data (i.e. void pixels).

property	NoData { get }
type	float64
value		The nodata value.
inherited	IImageInfo.NoData

Remarks:

The default value is NanD (i.e. all raw pixel values are valid).

See also:

IImageReader

PathInfo

Returns a Path object that represents the file path this object is associated with.

property	PathInfo { get }
type	Path
value	[not-null]	The path info.
inherited	IPathInfo.PathInfo

Remarks:

Objects that do not have a meaningful file path association simply return Unknown.

PixelFormat

The pixel format of the image.

property	PixelFormat { get }
type	PixelFormat
value		The pixel format of the image.
inherited	IImageInfo.PixelFormat

Remarks:

Image processing is always performed with with four 24-bit integer channels (see Scanline). The pixel format provides additional information about the stored pixel data.

Range

The mapping between raw pixels values and normalized image pixels.

property	Range { get }
type	PixelRange
value		The pixel range.
inherited	IImageInfo.Range

See also:

IImageReader

ScanlineY

The Y-coordinate that will be read upon the next call to ReadScanline.

property	ScanlineY { get }
type	int32
value	[0..Height]	The scanline Y-coordinate.

Width

Width of the image, in pixels.

property	Width { get }
type	int32
value	[>0]	Width of the image, in pixels.
inherited	IImageInfo.Width

Remarks:

This is the number of image pixels along the X-axis, which may be different from the raster width (see MapInfo), depending on the pixel coverage.

Methods

AcquireTry

Acquires a strong reference to this disposable object.

[OwnerReturn, ThreadSafe]
method	AcquireTry ()
type	IDisposable
returns			this if a new strong reference has been acquired, null if this object is already being disposed.
inherited	IDisposable.AcquireTry

Remarks:

The object will not be actually disposed by calls to Dispose when there is at least one strong reference left. Code that calls the AcquireTry method is responsible for calling the Dispose method accordingly.

This method is not intended to be used in performance-critical code. It should only be used to high-level resource management.

Adjust

Adjust the brightness and saturation of this image.

[OwnerReturn, OwnerThis]
method	Adjust (int32 brightness, int32 saturation = 0)
type	T
params	brightness	[-1000..1000]	The adjustment factor for image brightness.
	saturation	[-1000..1000]	The adjustment factor for image saturation. Defaults to 0.
returns		[not-null]	The resulting image.
inherited	IImageOps.Adjust

See also:

Colors.Adjust

Background

The image scanlines will be read using concurrent background tasks.

[OwnerReturn, OwnerThis]
method	Background ()
type	T
returns		[not-null]	The resulting IImageReader object.
inherited	IImageOps.Background

ChannelToIntensity

Returns an opaque grayscale image which pixel values are equal to the given pixel channel of this image.

[OwnerReturn, OwnerThis]
method	ChannelToIntensity (PixelChannel channel)
type	T
params	channel		The pixel channel.
returns		[not-null]	The resulting image.
inherited	IImageOps.ChannelToIntensity

Remarks:

Void pixels in this image will also be void in the resulting image.

See also:

IImageOps.IntensityToChannel

Clip

Returns an image that represents a sub rectangle of this image.

[OwnerReturn, OwnerThis]
method	Clip (int32 x, int32 y, int32 width, int32 height)
type	T
params	x	[>=0]	X-coordinate of top-left corner of rectangle.
	y	[>=0]	Y-coordinate of top-left corner of rectangle.
	width	[>0]	Width of rectangle.
	height	[>0]	Height of rectangle.
returns		[not-null]	The resulting image.
inherited	IImageOps.Clip

Crop

Assumes the image holds a non-rectangular shape that is padded with a uniform color and replaces all padding pixels with void.

[OwnerReturn, OwnerThis]
method	Crop (int64 color = Colors.Black, int32 skip = 1)
type	T
params	color		The padding pixel color (see Colors). Defaults to Black.
	skip		The maximum length (in pixels) of a run surrounded by void pixels that will be interpreted as void, too. Defaults to 1.
returns		[not-null]	The resulting image.
inherited	IImageOps.Crop

Remarks:

Since the source image is read scanline after scanline, cropping is performed locally for each scanline: Beginning at the left and right image borders, all pixels having the same color as the color parameter are replaced with void until a consecutive run of pixels having different colors is found (see skip). The following figures illustrates this: Pixels that have the same color as the color parameter are shown as '.' characters, all others are indicated with X:

+-----------------------------------------------------------------------+
|.......................................................................|
|.......................................................................|
|............................XXX........................................|
|...........................XXXXX.......................................|
|.........................XXXXXXX.......................................|
|.......................XXXXXXXXX.......................................|
|.....................XXXXXXXXXXX.......................................|
|...................XXXXXXXXXXXXXX......................................|
|.................XXXXXXXXXXXXXXXX......................................|
|................XXXXXXXXXXXXXXXXX......................................|
|...............XXXXXXXXXXXXXXXXXXX.....................................|
|................XXXXXXXXXXXXXXXXXXX....................................|
|.................XXXXXXXXXXXXXXXXXXXXX.................................|
|...................XXXXXXXXXXXXXXXXXXXXXXX.....X.......................|
|...............XXXXXXXXXX..XXXXXXXXXXXXXXXXXXXXXXXXXXX.................|
|..........XXXXXXXXXXXX.......XXXXXXXXXXXXXXXXXXXXXXXXXXX...............|
|.............XXXXXXX..........XXXXXXXXXXXXXXXXXXXXXXXXXX...............|
|...............XXX.............XXXXXXXXXXXXXXXXXXXXXXXX................|
|................................XXXXXXXXXXXXXXXXXXXXX..................|
|................................XXXXXXXXXXXXXXXXX......................|
|...................................XXXXXXX.............................|
|.......................................................................|
|.......................................................................|
|.......................................................................|
|.......................................................................|
+-----------------------------------------------------------------------+

The cropped image will look like this (whitespace represents void pixels):

+-----------------------------------------------------------------------+
|                                                                       |
|                                                                       |
|                            XXX                                        |
|                           XXXXX                                       |
|                         XXXXXXX                                       |
|                       XXXXXXXXX                                       |
|                     XXXXXXXXXXX                                       |
|                   XXXXXXXXXXXXXX                                      |
|                 XXXXXXXXXXXXXXXX                                      |
|                XXXXXXXXXXXXXXXXX              cropped because         |
|               XXXXXXXXXXXXXXXXXXX             of skipped pixels       |
|                XXXXXXXXXXXXXXXXXXX            |                       |
|                 XXXXXXXXXXXXXXXXXXXXX         V                       |
|                   XXXXXXXXXXXXXXXXXXXXXXX                             |
|               XXXXXXXXXX..XXXXXXXXXXXXXXXXXXXXXXXXXXX                 |
|          XXXXXXXXXXXX.......XXXXXXXXXXXXXXXXXXXXXXXXXXX               |
|             XXXXXXX..........XXXXXXXXXXXXXXXXXXXXXXXXXX               |
|               XXX.............XXXXXXXXXXXXXXXXXXXXXXXX                |
|                         ^      XXXXXXXXXXXXXXXXXXXXX                  |
|                         |      XXXXXXXXXXXXXXXXX                      |
|                  not reachable    XXXXXXX                             |
|                  from left or                                         |
|                  right border                                         |
|                                                                       |
|                                                                       |
+-----------------------------------------------------------------------+

Cropping will not change the image size or modify the georeference.

Dispose

Releases all resources held by this object if there are no more strong references to it, decrements the reference counter by one otherwise.

[Dispose, OwnerThis, ThreadSafe]
method	Dispose ()
inherited	IDisposable.Dispose

Remarks:

The Dispose method silently returns if the object has already been disposed.

Function

Transform linear pixel values using a custom function.

[OwnerReturn, OwnerThis]
method	Function (string function, PixelChannel channel = PixelChannel.Intensity)
type	T
params	function	[not-null]	A function expression (see GrammarRuleExpression) that consumes input linear pixel values in the range [0..1] and returns output linear pixel value in the range [0..1]. Output values are clamped (including infinities), not-a-number output values produce void pixels (see Void).
	channel		The pixel channel the function shall be applied to. If Void, no transformation will be performed. Defaults to Intensity.
returns		[not-null]	The resulting image or this if channel is Void.
inherited	IImageOps.Function

Remarks:

The given function is evaluated on linear pixel values, that is after raw pixels have been translated to normalized pixel values (see SetRange) and after gamma correction has been applied (see SetGamma). The expression may use all members of the default Tinman script (see Tinman). The following arguments are available to the function expression:

• raw
  The input pixel value (raw).
• lin
  The input pixel value (linear) in the range [0..1].
• A, B, C, D, E, F
  Coefficients that can be used to transform between raw and linear pixel values:

  lin = pow(raw * A + B, C)
  raw = pow(lin, D) * E + F
  D   = 1 / C
  E   = 1 / A
  F   = -B / A

  where raw is the raw pixel value that is actually stored in the source image (e.g. [0..255] for 8-bit images) and lin is the linear pixel value that is stored in the image scanline buffer (see Scanline). A and B represent the transformation from raw pixel values to normalized pixel values (see SetRange); C is the exponent that best approximates the performed gamma correction (see SetGamma).

IntensityToChannel

Returns an image that has a single pixel channel (see channel parameter), initialized to the grayscale intensities of this image. All other pixel channels are initialized to their respective default values (see PixelChannel).

[OwnerReturn, OwnerThis]
method	IntensityToChannel (PixelChannel channel)
type	T
params	channel		The pixel channel.
returns		[not-null]	The resulting image.
inherited	IImageOps.IntensityToChannel

Remarks:

Void pixels in this image will also be void in the resulting image.

See also:

IImageOps.ChannelToIntensity

---

Replaces a pixel channel of the given image with the grayscale intensities of this image.

[OwnerReturn, OwnerThis]
method	IntensityToChannel (PixelChannel channel, [Owner] T image)
type	T
params	channel		The pixel channel to replace.
	image	[not-null]	The image which pixel channel is to be replaced.
returns		[not-null]	The resulting image.
inherited	IImageOps.IntensityToChannel

Remarks:

Void pixels in the other image are ignored (i.e. the original pixel remains unchanged). Void pixels in this image are preserved (i.e. the other pixel is ignored).

See also:

PixelChannel

Invert

Inverts the given pixel channel.

[OwnerReturn, OwnerThis]
method	Invert (PixelChannel channel)
type	T
params	channel		The pixel channel.
returns		[not-null]	The resulting image.
inherited	IImageOps.Invert

Remarks:

The Void pixel channel cannot be inverted. The method will do nothing in this case.

Op

Combines this image (first) with the given one (second) using the specified color operation (see ColorOp).

[OwnerReturn, OwnerThis]
method	Op ([Owner] T other, ColorOp colorOp)
type	T
params	other	[not-null]	The other image.
	colorOp		The color operation to apply.
returns		[not-null]	The resulting image.
inherited	IImageOps.Op

Pad

Pads this image by duplicating pixels at the borders.

[OwnerReturn, OwnerThis]
method	Pad (int32 amount = 1)
type	T
params	amount	[>=0]	The pad amount, in pixels. Defaults to 1.
returns		[not-null]	The resulting image.
inherited	IImageOps.Pad

Remarks:

Padding can be used to reduce seams that will occur when tiled geodata is imported incrementally using on-the-fly projection. The correct solution would be to merge all geodata tiles into a dataset and then perform the projection on the whole geodata.

ReadAll

Reads the entire image into a color buffer.

[OwnerThis]
method	ReadAll ()
type	ColorBuffer
returns		[not-null]	The color buffer.

Remarks:

This method must only be called when the current scanline is zero (see ScanlineY).

ReadScanline

Reads the next scanline.

method	ReadScanline (Scanline scanline)
type	int32
params	scanline	[not-null]	The output buffer for the read scanline pixels.
returns			Y-coordinate of processed scanline.

Scale

Returns an image that represents a scaled version of this image.

[OwnerReturn, OwnerThis]
method	Scale (int32 width, int32 height, InterpolationMode mode = InterpolationMode.UniformB)
type	T
params	width	[>0]	Width of the scaled image.
	height	[>0]	Height of the scaled image.
	mode		The interpolation mode to use for magnification. Defaults to UniformB.
returns		[not-null]	The resulting image.
inherited	IImageOps.Scale

SetCoverage

Updates the value of Coverage, without modifying the image content.

[OwnerReturn, OwnerThis]
method	SetCoverage (PixelCoverage coverage)
type	T
params	coverage		The pixel coverage.
returns		[not-null]	The resulting image.
inherited	IImageOps.SetCoverage

Remarks:

To perform resampling between pixel-is-area and pixel-is point, use ToCoverage.

SetGamma

Updates the value of Gamma.

[OwnerReturn, OwnerThis]
method	SetGamma (float64 gamma)
type	T
params	gamma		The gamma correction.
returns		[not-null]	The resulting image.
inherited	IImageOps.SetGamma

SetGeoref

Updates the value of Georef, without modifying the image content.

[OwnerReturn, OwnerThis]
method	SetGeoref (Raster georef)
type	T
params	georef		The georeferencing information or null.
returns		[not-null]	The resulting image.
inherited	IImageOps.SetGeoref

SetLayer

Updates the value of Layer, without modifying the image content.

[OwnerReturn, OwnerThis]
method	SetLayer (HeightmapLayer layer, PixelFormat format = PixelFormat.Unknown)
type	T
params	layer		The heightmap layer.
	format		The pixel format to use for choosing the heightmap format (see For). If Unknown, the pixel format of this image will be used. Defaults to Unknown.
returns		[not-null]	The resulting image.
inherited	IImageOps.SetLayer

SetNoData

Updates the value of NoData.

[OwnerReturn, OwnerThis]
method	SetNoData (float64 noData)
type	T
params	noData		The raw pixel value to map to void pixels (see NoData).
returns		[not-null]	The resulting image.
inherited	IImageOps.SetNoData

SetRange

Updates the value of Range.

[OwnerReturn, OwnerThis]
method	SetRange (PixelRange range, bool reinterpret = false)
type	T
params	range		The pixel value ranges to use for translating raw pixel values to normalized pixel values (see Range).
	reinterpret		true to leave raw pixels values unchanged and reinterpret them according to range, false to normalize raw pixel values to range. Defaults to false.
returns		[not-null]	The resulting image.
inherited	IImageOps.SetRange

Sharpen

Performs sharpening on this image.

[OwnerReturn, OwnerThis]
method	Sharpen (IImage detail, float32 scale = 1, int32 downsample = 4)
type	T
params	detail	[not-null]	The high-resolution detail image to use for sharpening.
	scale	[>=0]	Scale factor to apply to difference between full and reduced resolution. Defaults to 1.
	downsample	[>0]	The detail image will be downsamples by this factor. The difference between the full and reduced resolutions will be used to sharpen this image. Defaults to 4.
returns		[not-null]	The resulting image.
inherited	IImageOps.Sharpen

SkipScanlines

Skips the given number of scanlines.

method	SkipScanlines (int32 count)
params	count	[>=0]	The number of scanlines to skip.

Smooth

Performs smoothing on this image.

[OwnerReturn, OwnerThis]
method	Smooth ()
type	T
returns		[not-null]	The resulting image.
inherited	IImageOps.Smooth

Remarks:

Image smoothing is performed using an Uniform-B spline.

ToConfig

Returns the configuration value that describes this object.

method	ToConfig ()
type	ConfigValue
returns		[not-null]	The configuration value.
inherited	IConfigurable.ToConfig

Remarks:

All configurable objects need to implement this interface. For simple types, it is preferable to delegate to ToValue.

The returned value may be of type Invalid, which means that this object in its current state cannot be described with the configuration API.

Before returning the resulting configuration value, Cache must be called on it, passing this IConfigurable object as parameter.

ToCoverage

Transforms this image from the given pixel coverage semantic to the given one.

[OwnerReturn, OwnerThis]
method	ToCoverage (PixelCoverage target, bool wrapX = false)
type	T
params	target		The pixel coverage semantic of the resulting image.
	wrapX		Is this image wrapped around the X-axis?
returns		[not-null]	The resulting image.
inherited	IImageOps.ToCoverage

Remarks:

If this image already has the given pixel coverage semantic, no transformation will be performed.

Validate

Validates the state of this object.

method	Validate (Validator validator)
params	validator	[not-null]	The validator object.
inherited	IValidatable.Validate