org.esa.beam.framework.gpf
Interface Tile

All Superinterfaces:

Iterable<Tile.Pos>

All Known Implementing Classes:

TileImpl

public interface Tile
extends Iterable<Tile.Pos>

A tile represents a rectangular region of sample data within the scene rectangle of a data product. Tiles are used to enable the sample data transfer from and to the source and target bands of data products used within operator graphs.

Target tiles to be computed are passed into an Operator's computeTile and computeTileStack computeTileStack} methods. Source tiles are obtained by using the getSourceTile method.

Three ways are provided to access and manipulate the sample data of a target tile:

(1) This is the simplest (but also slowest) way to modify sample data of a tile:

   for (int y = tile.getMinY(); y <= tile.getMaxY(); y++) {
       for (int x = tile.getMinX(); x <= tile.getMaxX(); x++) {
           // compute sample value...
           tile.setSample(x, y, sample);
       }
   }
 

which can also be written even simpler using a tile iterator:

   for (Tile.Pos pos : tile) {
       // compute sample value...
       tile.setSample(pos.x, pos.y, sample);
   }
 

The methods getSampleFloat(int, int) and setSample(int, int, float) and their derivatives used in option (1) return and expect (geo-)physically scaled sample values.

(2) More performance is gained if the sample data buffer is checked out and committed (required after modification only):

   ProductData samples = tile.getRawSamples(); // check out
   for (int y = 0; y < tile.getHeight(); y++) {
       for (int x = 0; x < tile.getWidth(); x++) {
           // compute sample value...
           samples.setElemFloatAt(y * getWidth() + x, sample);
           // ...
       }
   }
   tile.setRawSamples(samples); // commit
 

The method getRawSamples() used in option (2) returns a writable buffer for the raw, non-calibrated sample values. Use the toGeoPhysical(float) and toRaw(float) to convert between physical and raw sample values.

(3) The the fastest way to read from or write to sample data is to directly access the sample data via their primitive data buffers:

   float[] samples = tile.getDataBufferFloat();
   float sample;
   int offset = tile.getScanlineOffset();
   for (int y = 0; y < tile.getHeight(); y++) {
       int index = offset;
       for (int x = 0; x < tile.getWidth(); x++) {
           // compute sample value...
           samples[index] = sample;
           index++;
       }
       offset += tile.getScanlineStride();
   }
 

Note that option (3) can only be used if the exact sample data type is known or has been identified in a former step. The code snippet above implies that the underlying data type is float (because getRasterDataNode().getDataType() returns ProductData.TYPE_FLOAT32).
The getDataBufferFloat() and its derivatives all return arrays of raw, non-calibrated sample values. Use the toGeoPhysical(float) and toRaw(float) to convert between physical and raw sample values.

Since:

4.1

Author:

Norman Fomferra, Marco Peters, Marco Zühlke

Nested Class Summary

static class

Tile.Pos A pixel position within the tile's raster.

Method Summary

ProductData	getDataBuffer() Obtains access to the underlying raw sample buffer.
byte[]	getDataBufferByte() Gets the underlying raw sample array of type byte (signed or unsigned).
double[]	getDataBufferDouble() Gets the underlying raw sample array of type double.
float[]	getDataBufferFloat() Gets the underlying raw sample array of type float.
int	getDataBufferIndex(int x, int y) Gets the index into the underlying raw sample buffer for the given pixel coordinates.
int[]	getDataBufferInt() Gets the underlying raw sample array of type int.
short[]	getDataBufferShort() Gets the underlying raw sample array of type short (signed or unsigned).
int	getHeight() Gets the height in pixels within the scene covered by the tile's RasterDataNode.
int	getMaxX() Gets the maximum pixel x-coordinate within the scene covered by the tile's RasterDataNode.
int	getMaxY() Gets the maximum pixel y-coordinate within the scene covered by the tile's RasterDataNode.
int	getMinX() Gets the minimum pixel x-coordinate within the scene covered by the tile's RasterDataNode.
int	getMinY() Gets the minimum pixel y-coordinate within the scene covered by the tile's RasterDataNode.
RasterDataNode	getRasterDataNode() Gets the RasterDataNode associated with this tile, e.g.
ProductData	getRawSamples() Gets the raw (unscaled, uncalibrated) samples, e.g.
Rectangle	getRectangle() Gets the tile rectangle in pixel coordinates within the scene covered by the tile's RasterDataNode.
boolean	getSampleBit(int x, int y, int bitIndex) Gets the bit-coded sample value for the given pixel coordinate and the specified bit index as a boolean.
boolean	getSampleBoolean(int x, int y) Gets the (geo-)physically scaled sample at the given pixel coordinate as boolean value.
double	getSampleDouble(int x, int y) Gets the (geo-)physically scaled sample value for the given pixel coordinate as double.
float	getSampleFloat(int x, int y) Gets the (geo-)physically scaled sample at the given pixel coordinate as float value.
int	getSampleInt(int x, int y) Gets the (geo-)physically scaled sample at the given pixel coordinate as int value.
double[]	getSamplesDouble() Gets the scaled, (geo-)physical array of double samples, copied from or directly returning the underlying data buffer.
float[]	getSamplesFloat() Gets the scaled, (geo-)physical array of double samples, copied from or directly returning the underlying data buffer.
int[]	getSamplesInt() Gets the scaled, (geo-)physical array of int samples, copied from or directly returning the underlying data buffer.
int	getScanlineOffset() Gets the scanline offset.
int	getScanlineStride() Gets the raster scanline stride for addressing the internal data buffer.
int	getWidth() Gets the width in pixels within the scene covered by the tile's RasterDataNode.
boolean	isSampleValid(int x, int y) Checks whether or not the sample value exists and is valid at a given image pixel position.
boolean	isTarget() Checks if this is a target tile.
Iterator<Tile.Pos>	iterator() Gets an iterator which can be used to visit all pixels in the tile.
void	setRawSamples(ProductData rawSamples) Sets this tile's raw (unscaled, uncalibrated) samples.
void	setSample(int x, int y, boolean sample) Sets the (geo-)physically scaled sample at the given pixel coordinate from a boolean value.
void	setSample(int x, int y, double sample) Sets the (geo-)physically scaled sample at the given pixel coordinate from a double value.
void	setSample(int x, int y, float sample) Sets the (geo-)physically scaled sample at the given pixel coordinate from a float value.
void	setSample(int x, int y, int sample) Sets the (geo-)physically scaled sample at the given pixel coordinate from a int value.
void	setSample(int x, int y, int bitIndex, boolean sample) Sets the bit-coded sample at the given pixel coordinate and the specified bit index from a boolean value.
void	setSamples(double[] samples) Sets this tile's scaled, (geo-)physical samples as array of doubles.
void	setSamples(float[] samples) Sets this tile's scaled, (geo-)physical samples as array of floatss.
void	setSamples(int[] samples) Sets this tile's scaled, (geo-)physical samples as array of floatss.
double	toGeoPhysical(double rawSample) Converts a raw sample value (e.g.
float	toGeoPhysical(float rawSample) Converts a raw sample value (e.g.
double	toRaw(double sample) Converts a (geo-)physically scaled sample value of type double to a its corresponding raw sample value (e.g.
float	toRaw(float sample) Converts a (geo-)physically scaled sample value of type float to a its corresponding raw sample value (e.g.

Method Detail

getRasterDataNode

RasterDataNode getRasterDataNode()

Gets the RasterDataNode associated with this tile, e.g. a Band for source and target tiles or TiePointGrid for a source tile.

Returns:

The RasterDataNode associated with this tile.

isTarget

boolean isTarget()

Checks if this is a target tile. Non-target tiles are read only.

Returns:

true if this is a target tile.

toGeoPhysical

float toGeoPhysical(float rawSample)

Converts a raw sample value (e.g. digital counts) to a (geo-)physically scaled sample value of type float.

Parameters:

rawSample - The raw sample value.

Returns:

The calibrated sample value.

toGeoPhysical

double toGeoPhysical(double rawSample)

Converts a raw sample value (e.g. digital counts) to a (geo-)physically scaled sample value of type double.

Parameters:

rawSample - The raw sample value.

Returns:

The calibrated sample value.

toRaw

float toRaw(float sample)

Converts a (geo-)physically scaled sample value of type float to a its corresponding raw sample value (e.g. digital counts).

Parameters:

sample - The calibrated sample value.

Returns:

The raw sample value.

toRaw

double toRaw(double sample)

Converts a (geo-)physically scaled sample value of type double to a its corresponding raw sample value (e.g. digital counts).

Parameters:

sample - The calibrated sample value.

Returns:

The raw sample value.

getRectangle

Rectangle getRectangle()

Gets the tile rectangle in pixel coordinates within the scene covered by the tile's RasterDataNode. Simply returns new Rectangle( minX, minY, width, height).

Returns:

The tile rectangle in pixel coordinates.

getMinX

int getMinX()

Gets the minimum pixel x-coordinate within the scene covered by the tile's RasterDataNode.

Returns:

The minimum pixel x-coordinate.

getMaxX

int getMaxX()

Gets the maximum pixel x-coordinate within the scene covered by the tile's RasterDataNode.

Returns:

The maximum pixel x-coordinate.

getMinY

int getMinY()

Gets the minimum pixel y-coordinate within the scene covered by the tile's RasterDataNode.

Returns:

The minimum pixel y-coordinate.

getMaxY

int getMaxY()

Gets the maximum pixel y-coordinate within the scene covered by the tile's RasterDataNode.

Returns:

The maximum pixel y-coordinate.

getWidth

int getWidth()

Gets the width in pixels within the scene covered by the tile's RasterDataNode.

Returns:

The width in pixels.

getHeight

int getHeight()

Gets the height in pixels within the scene covered by the tile's RasterDataNode.

Returns:

The height in pixels.

getDataBufferIndex

int getDataBufferIndex(int x,
                       int y)

Gets the index into the underlying raw sample buffer for the given pixel coordinates.

The pixel coordinates are absolute; meaning they are defined in the scene raster coordinate system of this tile's RasterDataNode.

The returned index is computed as follows:

   int dx = x - getMinX();
   int dy = y - getMinY();
   int index = getScanlineOffset() + dy * getScanlineStride() + dx;
 

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

Returns:

an index into the underlying data buffer

getDataBuffer

ProductData getDataBuffer()

Obtains access to the underlying raw sample buffer. The data buffer holds the raw (unscaled, uncalibrated) sample data (e.g. detector counts).

Elements in this array must be addressed by an index computed via the scanlineStride and scanlineOffset properties. The index can also be directly computed using the getDataBufferIndex(int, int) method.

The most efficient way to access and/or modify the samples in the raw data buffer is using the following nested loops:

   int lineStride = tile.getScanlineStride();
   int lineOffset = tile.getScanlineOffset();
   for (int y = tile.getMinY(); y <= tile.getMaxY(); y++) {
      int index = lineOffset;
      for (int x = tile.getMinX(); x <= tile.getMaxX(); x++) {
           // use index here to access raw data buffer...
           index++;
       }
       lineOffset += lineStride;
   }
 

If the absolute x,y pixel coordinates are not required, the following construct maybe more readable:

   int lineStride = tile.getScanlineStride();
   int lineOffset = tile.getScanlineOffset();
   for (int y = 0; y < tile.getHeight(); y++) {
      int index = lineOffset;
      for (int x = 0; x < tile.getWidth(); x++) {
           // use index here to access raw data buffer...
           index++;
       }
       lineOffset += lineStride;
   }
 

Returns:

the sample data

getDataBufferByte

byte[] getDataBufferByte()

Gets the underlying raw sample array of type byte (signed or unsigned). If the underlying data buffer is not of type byte, null is returned.

Refer to getDataBuffer() for using the primitive array.

Returns:

The underlying data buffer's primitive array, or null (see above).

See Also:

getDataBufferIndex(int, int), getDataBuffer()

getDataBufferShort

short[] getDataBufferShort()

Gets the underlying raw sample array of type short (signed or unsigned). If the underlying data buffer is not of type short, null is returned.

Refer to getDataBuffer() for using the primitive array.

Returns:

The underlying data buffer's primitive array, or null (see above).

See Also:

getDataBufferIndex(int, int), getDataBuffer()

getDataBufferInt

int[] getDataBufferInt()

Gets the underlying raw sample array of type int. If the underlying data buffer is not of type int, null is returned.

Refer to getDataBuffer() for using the primitive array.

Returns:

The underlying data buffer's primitive array, or null (see above).

See Also:

getDataBufferIndex(int, int), getDataBuffer()

getDataBufferFloat

float[] getDataBufferFloat()

Gets the underlying raw sample array of type float. If the underlying data buffer is not of type float, null is returned.

Refer to getDataBuffer() for using the primitive array.

Returns:

The underlying data buffer's primitive array, or null (see above).

See Also:

getDataBufferIndex(int, int), getDataBuffer()

getDataBufferDouble

double[] getDataBufferDouble()

Gets the underlying raw sample array of type double. If the underlying data buffer is not of type double, null is returned.

Refer to getDataBuffer() for using the primitive array.

Returns:

The underlying data buffer's primitive array, or null (see above).

See Also:

getDataBufferIndex(int, int), getDataBuffer()

getScanlineOffset

int getScanlineOffset()

Gets the scanline offset. The scanline offset is the index to the first valid sample element in the data buffer.

Returns:

The raster scanline offset.

See Also:

getScanlineStride()

getScanlineStride

int getScanlineStride()

Gets the raster scanline stride for addressing the internal data buffer. The scanline stride is added to the scanline offset in order to compute offsets of subsequent scanlines.

Returns:

The raster scanline stride.

See Also:

getScanlineOffset()

getRawSamples

ProductData getRawSamples()

Gets the raw (unscaled, uncalibrated) samples, e.g. detector counts, copied from or wrapping the underlying data buffer. In contradiction to the getDataBuffer() method, the returned samples will cover exactly the region getRectangle() rectangle} of this tile. Thus, the number of returned samples will always equal width * height.

In order to apply changes of the samples values to this tile, it is mandatory to call setRawSamples(org.esa.beam.framework.datamodel.ProductData) with the modified ProductData instance.

Returns:

The raw samples copied from or wrapping the underlying data buffer.

setRawSamples

void setRawSamples(ProductData rawSamples)

Sets this tile's raw (unscaled, uncalibrated) samples. The number of given samples must be equal width * height of this tile.

This method must be used in order to apply changes made to the samples returned by the getRawSamples() method.

Parameters:

rawSamples - The raw samples to be set.

getSamplesInt

int[] getSamplesInt()

Gets the scaled, (geo-)physical array of int samples, copied from or directly returning the underlying data buffer. In contradiction to the getDataBuffer() method, the returned samples will cover exactly the region getRectangle() rectangle} of this tile. Thus, the number of returned samples will always equal width * height.

Sample values that are masked out (see isSampleValid(int, int)) are returned as array elements having the value Float.NaN.

Returns:

The (geo-)physical samples computed from the underlying raw data buffer.

Since:

BEAM 4.8

See Also:

setSamples(int[])

getSamplesFloat

float[] getSamplesFloat()

Gets the scaled, (geo-)physical array of double samples, copied from or directly returning the underlying data buffer. In contradiction to the getDataBuffer() method, the returned samples will cover exactly the region getRectangle() rectangle} of this tile. Thus, the number of returned samples will always equal width * height.

Sample values that are masked out (see isSampleValid(int, int)) are returned as array elements having the value Float.NaN.

Returns:

The (geo-)physical samples computed from the underlying raw data buffer.

See Also:

setSamples(float[])

getSamplesDouble

double[] getSamplesDouble()

Gets the scaled, (geo-)physical array of double samples, copied from or directly returning the underlying data buffer. In contradiction to the getDataBuffer() method, the returned samples will cover exactly the region getRectangle() rectangle} of this tile. Thus, the number of returned samples will always equal width * height.

Sample values that are masked out (see isSampleValid(int, int)) are returned as array elements having the value Float.NaN.

Returns:

The (geo-)physical samples computed from the underlying raw data buffer.

See Also:

setSamples(double[])

setSamples

void setSamples(int[] samples)

Sets this tile's scaled, (geo-)physical samples as array of floatss. The number of given samples must be equal width * height of this tile.

Parameters:

samples - The (geo-)physical samples to be set.

Since:

BEAM 4.8

See Also:

getSamplesInt()

setSamples

void setSamples(float[] samples)

Sets this tile's scaled, (geo-)physical samples as array of floatss. The number of given samples must be equal width * height of this tile.

Parameters:

samples - The (geo-)physical samples to be set.

See Also:

getSamplesFloat()

setSamples

void setSamples(double[] samples)

Sets this tile's scaled, (geo-)physical samples as array of doubles. The number of given samples must be equal width * height of this tile.

This method must be used in order to apply changes made to the samples returned by the getRawSamples() method.

Parameters:

samples - The (geo-)physical samples to be set.

See Also:

getSamplesDouble()

isSampleValid

boolean isSampleValid(int x,
                      int y)

Checks whether or not the sample value exists and is valid at a given image pixel position.

Parameters:

x - the image pixel x-coordinate

y - the image pixel y-coordinate

Returns:

true, if the sample exists and is valid

Since:

BEAM 4.7.1

getSampleBoolean

boolean getSampleBoolean(int x,
                         int y)

Gets the (geo-)physically scaled sample at the given pixel coordinate as boolean value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

Returns:

The (geo-)physical sample as boolean value.

setSample

void setSample(int x,
               int y,
               boolean sample)

Sets the (geo-)physically scaled sample at the given pixel coordinate from a boolean value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

sample - The (geo-)physically scaled sample as boolean value.

getSampleInt

int getSampleInt(int x,
                 int y)

Gets the (geo-)physically scaled sample at the given pixel coordinate as int value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

Returns:

The (geo-)physically scaled sample as int value.

setSample

void setSample(int x,
               int y,
               int sample)

Sets the (geo-)physically scaled sample at the given pixel coordinate from a int value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed. The conversion ensures that no overflow happens. If necessary the value is cropped to the value range.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

sample - The (geo-)physically scaled sample as int value.

getSampleFloat

float getSampleFloat(int x,
                     int y)

Gets the (geo-)physically scaled sample at the given pixel coordinate as float value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases it is more performant to directly access the tile's dataBuffer in conjunction with the scanlineOffset and scanlineStride properties.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

Returns:

The (geo-)physically scaled sample as float value.

setSample

void setSample(int x,
               int y,
               float sample)

Sets the (geo-)physically scaled sample at the given pixel coordinate from a float value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed. The conversion ensures that no overflow happens. If necessary the value is cropped to the value range.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

sample - The (geo-)physically scaled sample as float value.

getSampleDouble

double getSampleDouble(int x,
                       int y)

Gets the (geo-)physically scaled sample value for the given pixel coordinate as double.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

Returns:

The (geo-)physically scaled sample as double value.

setSample

void setSample(int x,
               int y,
               double sample)

Sets the (geo-)physically scaled sample at the given pixel coordinate from a double value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed. The conversion ensures that no overflow happens. If necessary the value is cropped to the value range.

Note that in most cases, accessing the tile's dataBuffer directly in conjunction with the scanlineOffset and scanlineStride properties gives a better performance.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

sample - The (geo-)physically scaled sample as double value.

getSampleBit

boolean getSampleBit(int x,
                     int y,
                     int bitIndex)

Gets the bit-coded sample value for the given pixel coordinate and the specified bit index as a boolean.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases it is more performant to directly access the tile's dataBuffer in conjunction with the scanlineOffset and scanlineStride properties.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

bitIndex - The bit index.

Returns:

The sample as boolean value.

setSample

void setSample(int x,
               int y,
               int bitIndex,
               boolean sample)

Sets the bit-coded sample at the given pixel coordinate and the specified bit index from a boolean value.

If the underlying data buffer is of a different sample data type, an appropriate type conversion is performed.

Note that in most cases it is more performant to directly access the tile's dataBuffer in conjunction with the scanlineOffset and scanlineStride properties.

Parameters:

x - The absolute pixel x-coordinate, must be greater or equal getMinX() and less or equal getMaxX().

y - The absolute pixel y-coordinate, must be greater or equal getMinY() and less or equal getMaxY().

bitIndex - The bit index.

sample - The sample as boolean value.

iterator

Iterator<Tile.Pos> iterator()

Gets an iterator which can be used to visit all pixels in the tile. The method allows this tile to be the target of the Java "foreach" statement. Using the tile as an iterator in a single loop

 for (Tile.Pos pos: tile) {
    int x = pos.x;
    int y = pos.y;
    // ...
 }
 

is equivalent to iterating over all pixels using two nested loops

 for (int y = tile.getMinY(); y <= tile.getMaxY(); y++) {
     for (int x = tile.getMinX(); x <= tile.getMaxX(); x++) {
         // ...
     }
 }
 

Specified by:

iterator in interface Iterable<Tile.Pos>