Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
package nom.tam.fits.compression.algorithm.quant;
/*
* #%L
* nom.tam FITS library
* %%
* Copyright (C) 1996 - 2024 nom-tam-fits
* %%
* This is free and unencumbered software released into the public domain.
*
* Anyone is free to copy, modify, publish, use, compile, sell, or
* distribute this software, either in source code form or as a compiled
* binary, for any purpose, commercial or non-commercial, and by any
* means.
*
* In jurisdictions that recognize copyright laws, the author or authors
* of this software dedicate any and all copyright interest in the
* software to the public domain. We make this dedication for the benefit
* of the public at large and to the detriment of our heirs and
* successors. We intend this dedication to be an overt act of
* relinquishment in perpetuity of all present and future rights to this
* software under copyright law.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
* #L%
*/
import java.nio.ByteBuffer;
import java.nio.DoubleBuffer;
import java.nio.FloatBuffer;
import java.nio.IntBuffer;
import nom.tam.fits.compression.algorithm.api.ICompressor;
/**
* (for internal use) Qunatization step processor as part of compression.
*/
@SuppressWarnings({"javadoc", "deprecation"})
public class QuantizeProcessor {
public static class DoubleQuantCompressor extends QuantizeProcessor implements ICompressor {
private final ICompressor postCompressor;
public DoubleQuantCompressor(QuantizeOption quantizeOption, ICompressor compressor) {
super(quantizeOption);
postCompressor = compressor;
}
@Override
public boolean compress(DoubleBuffer buffer, ByteBuffer compressed) {
IntBuffer intData = IntBuffer.wrap(new int[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()]);
double[] doubles = new double[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()];
buffer.get(doubles);
if (!this.quantize(doubles, intData)) {
return false;
}
intData.rewind();
postCompressor.compress(intData, compressed);
return true;
}
@Override
public void decompress(ByteBuffer compressed, DoubleBuffer buffer) {
IntBuffer intData = IntBuffer.wrap(new int[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()]);
postCompressor.decompress(compressed, intData);
intData.rewind();
unquantize(intData, buffer);
}
}
/**
* TODO this is done very inefficient and should be refactored!
*/
public static class FloatQuantCompressor extends QuantizeProcessor implements ICompressor {
private final ICompressor postCompressor;
public FloatQuantCompressor(QuantizeOption quantizeOption, ICompressor postCompressor) {
super(quantizeOption);
this.postCompressor = postCompressor;
}
@Override
public boolean compress(FloatBuffer buffer, ByteBuffer compressed) {
float[] floats = new float[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()];
double[] doubles = new double[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()];
buffer.get(floats);
for (int index = 0; index < doubles.length; index++) {
doubles[index] = floats[index];
}
IntBuffer intData = IntBuffer.wrap(new int[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()]);
if (!this.quantize(doubles, intData)) {
return false;
}
intData.rewind();
postCompressor.compress(intData, compressed);
return true;
}
@Override
public void decompress(ByteBuffer compressed, FloatBuffer buffer) {
IntBuffer intData = IntBuffer.wrap(new int[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()]);
postCompressor.decompress(compressed, intData);
intData.rewind();
double[] doubles = new double[quantizeOption.getTileHeight() * quantizeOption.getTileWidth()];
DoubleBuffer doubleBuffer = DoubleBuffer.wrap(doubles);
unquantize(intData, doubleBuffer);
for (double d : doubles) {
buffer.put((float) d);
}
}
}
private class BaseFilter extends PixelFilter {
BaseFilter() {
super(null);
}
@Override
protected void nextPixel() {
}
@Override
protected double toDouble(int pixel) {
return (pixel + ROUNDING_HALF) * bScale + bZero;
}
@Override
protected int toInt(double pixel) {
return nint((pixel - bZero) / bScale + ROUNDING_HALF);
}
}
private class DitherFilter extends PixelFilter {
private static final int RANDOM_MULTIPLICATOR = 500;
private int iseed = 0;
private int nextRandom = 0;
DitherFilter(long seed) {
super(null);
initialize(seed);
}
public void initialize(long ditherSeed) {
iseed = (int) ((ditherSeed - 1) % RandomSequence.length());
initI1();
}
private void initI1() {
nextRandom = (int) (RandomSequence.get(iseed) * RANDOM_MULTIPLICATOR);
}
public double nextRandom() {
return RandomSequence.get(nextRandom);
}
@Override
protected void nextPixel() {
nextRandom++;
if (nextRandom >= RandomSequence.length()) {
iseed++;
if (iseed >= RandomSequence.length()) {
iseed = 0;
}
initI1();
}
}
@Override
protected double toDouble(int pixel) {
return (pixel - nextRandom() + ROUNDING_HALF) * bScale + bZero;
}
@Override
protected int toInt(double pixel) {
return nint((pixel - bZero) / bScale + nextRandom() - ROUNDING_HALF);
}
}
private class NullFilter extends PixelFilter {
private final double nullValue;
private final boolean isNaN;
private final int nullValueIndicator;
NullFilter(double nullValue, int nullValueIndicator, PixelFilter next) {
super(next);
this.nullValue = nullValue;
isNaN = Double.isNaN(this.nullValue);
this.nullValueIndicator = nullValueIndicator;
}
public final boolean isNull(double pixel) {
return isNaN ? Double.isNaN(pixel) : nullValue == pixel;
}
@Override
protected double toDouble(int pixel) {
if (pixel == nullValueIndicator) {
return nullValue;
}
return super.toDouble(pixel);
}
@Override
protected int toInt(double pixel) {
if (isNull(pixel)) {
return nullValueIndicator;
}
return super.toInt(pixel);
}
}
private class PixelFilter {
private final PixelFilter next;
protected PixelFilter(PixelFilter next) {
this.next = next;
}
protected void nextPixel() {
next.nextPixel();
}
protected double toDouble(int pixel) {
return next.toDouble(pixel);
}
protected int toInt(double pixel) {
return next.toInt(pixel);
}
}
private class ZeroFilter extends PixelFilter {
ZeroFilter(PixelFilter next) {
super(next);
}
@Override
protected double toDouble(int pixel) {
if (pixel == ZERO_VALUE) {
return 0.0;
}
return super.toDouble(pixel);
}
@Override
protected int toInt(double pixel) {
if (pixel == 0.0) {
return ZERO_VALUE;
}
return super.toInt(pixel);
}
}
private static final double MAX_INT_AS_DOUBLE = Integer.MAX_VALUE;
/**
* number of reserved values, starting with
*/
private static final long N_RESERVED_VALUES = 10;
private static final double ROUNDING_HALF = 0.5;
/**
* value used to represent zero-valued pixels
*/
private static final int ZERO_VALUE = Integer.MIN_VALUE + 2;
private final boolean centerOnZero;
private final PixelFilter pixelFilter;
private double bScale;
private double bZero;
private Quantize quantize;
protected final QuantizeOption quantizeOption;
public QuantizeProcessor(QuantizeOption quantizeOption) {
this.quantizeOption = quantizeOption;
bScale = quantizeOption.getBScale();
bZero = quantizeOption.getBZero();
PixelFilter filter = null;
boolean localCenterOnZero = quantizeOption.isCenterOnZero();
if (quantizeOption.isDither2()) {
filter = new DitherFilter(quantizeOption.getSeed() + quantizeOption.getTileIndex());
localCenterOnZero = true;
quantizeOption.setCheckZero(true);
} else if (quantizeOption.isDither()) {
filter = new DitherFilter(quantizeOption.getSeed() + quantizeOption.getTileIndex());
} else {
filter = new BaseFilter();
}
if (quantizeOption.isCheckZero()) {
filter = new ZeroFilter(filter);
}
if (quantizeOption.isCheckNull()) {
final NullFilter nullFilter = new NullFilter(quantizeOption.getNullValue(), quantizeOption.getBNull(), filter);
filter = nullFilter;
quantize = new Quantize(quantizeOption) {
@Override
protected int findNextValidPixelWithNullCheck(int nx, DoubleArrayPointer rowpix, int ii) {
while (ii < nx && nullFilter.isNull(rowpix.get(ii))) {
ii++;
}
return ii;
}
@Override
protected boolean isNull(double d) {
return nullFilter.isNull(d);
}
};
} else {
quantize = new Quantize(quantizeOption);
}
pixelFilter = filter;
centerOnZero = localCenterOnZero;
}
public Quantize getQuantize() {
return quantize;
}
public boolean quantize(double[] doubles, IntBuffer quants) {
boolean success = quantize.quantize(doubles, quantizeOption.getTileWidth(), quantizeOption.getTileHeight());
if (success) {
calculateBZeroAndBscale();
quantize(DoubleBuffer.wrap(doubles, 0, quantizeOption.getTileWidth() * quantizeOption.getTileHeight()), quants);
}
return success;
}
public void quantize(final DoubleBuffer fdata, final IntBuffer intData) {
while (fdata.hasRemaining()) {
intData.put(pixelFilter.toInt(fdata.get()));
pixelFilter.nextPixel();
}
}
public void unquantize(final IntBuffer intData, final DoubleBuffer fdata) {
while (fdata.hasRemaining()) {
fdata.put(pixelFilter.toDouble(intData.get()));
pixelFilter.nextPixel();
}
}
private void calculateBZeroAndBscale() {
bScale = quantizeOption.getBScale();
bZero = zeroCenter();
quantizeOption.setIntMinValue(nint((quantizeOption.getMinValue() - bZero) / bScale));
quantizeOption.setIntMaxValue(nint((quantizeOption.getMaxValue() - bZero) / bScale));
quantizeOption.setBZero(bZero);
}
private int nint(double x) {
return x >= 0. ? (int) (x + ROUNDING_HALF) : (int) (x - ROUNDING_HALF);
}
private double zeroCenter() {
final double minValue = quantizeOption.getMinValue();
final double maxValue = quantizeOption.getMaxValue();
double evaluatedBZero;
if (!quantizeOption.isCheckNull() && !centerOnZero) {
// don't have to check for nulls
// return all positive values, if possible since some compression
// algorithms either only work for positive integers, or are more
// efficient.
if ((maxValue - minValue) / bScale < MAX_INT_AS_DOUBLE - N_RESERVED_VALUES) {
evaluatedBZero = minValue;
// fudge the zero point so it is an integer multiple of bScale
// This helps to ensure the same scaling will be performed if
// the file undergoes multiple fpack/funpack cycles
long iqfactor = (long) (evaluatedBZero / bScale + ROUNDING_HALF);
evaluatedBZero = iqfactor * bScale;
} else {
/* center the quantized levels around zero */
evaluatedBZero = (minValue + maxValue) / 2.;
}
} else {
// data contains null values or has be forced to center on zero
// shift the range to be close to the value used to represent null
// values
evaluatedBZero = minValue - bScale * (Integer.MIN_VALUE + N_RESERVED_VALUES + 1);
}
return evaluatedBZero;
}
}
