boofcv.alg.feature.orientation.impl.ImplOrientationImageAverageIntegral Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of boofcv-feature Show documentation
BoofCV is an open source Java library for real-time computer vision and robotics applications.
The newest version!
/*
 * Copyright (c) 2021, Peter Abeles. All Rights Reserved.
 *
 * This file is part of BoofCV (http://boofcv.org).
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package boofcv.alg.feature.orientation.impl;

import boofcv.abst.feature.orientation.RegionOrientation;
import boofcv.alg.feature.orientation.OrientationIntegralBase;
import boofcv.factory.transform.ii.FactorySparseIntegralFilters;
import boofcv.struct.convolve.Kernel2D_F64;
import boofcv.struct.image.ImageGray;
import boofcv.struct.sparse.GradientValue;
import boofcv.struct.sparse.SparseScaleSample_F64;

/**
 * 
 * Estimates the orientation of a region using a "derivative free" method. Points are sampled using
 * an integral image.
 * 
 *
 * @author Peter Abeles
 */
public class ImplOrientationImageAverageIntegral, G extends GradientValue>
		extends OrientationIntegralBase {
	// cosine values for each pixel
	protected Kernel2D_F64 kerCosine;
	// sine values for each pixel
	protected Kernel2D_F64 kerSine;

	private final SparseScaleSample_F64 sampler;

	/**
	 * @param sampleRadius Radius of the region being considered in terms of Wavelet samples. Typically 6.
	 * @param weightSigma Sigma for weighting distribution. Zero for unweighted.
	 */
	public ImplOrientationImageAverageIntegral( double radiusToScale,
												int sampleRadius, double period,
												int kernelWidth, double weightSigma,
												Class imageType ) {
		super(radiusToScale, sampleRadius, period, kernelWidth, weightSigma, false, imageType);

		int w = sampleRadius*2 + 1;
		kerCosine = new Kernel2D_F64(w);
		kerSine = new Kernel2D_F64(w);

		for (int y = -sampleRadius; y <= sampleRadius; y++) {
			int pixelY = y + sampleRadius;
			for (int x = -sampleRadius; x <= sampleRadius; x++) {
				int pixelX = x + sampleRadius;
				float r = (float)Math.sqrt(x*x + y*y);
				kerCosine.set(pixelX, pixelY, (float)x/r);
				kerSine.set(pixelX, pixelY, (float)y/r);
			}
		}
		kerCosine.set(sampleRadius, sampleRadius, 0);
		kerSine.set(sampleRadius, sampleRadius, 0);

		sampler = FactorySparseIntegralFilters.sample(imageType);
		setObjectRadius(1.0/objectRadiusToScale);
	}

	@Override
	public void setImage( T integralImage ) {
		super.setImage(integralImage);
		sampler.setImage(integralImage);
	}

	@Override
	public void setObjectRadius( double radius ) {
		super.setObjectRadius(radius);
		sampler.setWidth(kernelWidth*scale);
	}

	@Override
	public double compute( double c_x, double c_y ) {

		double period = scale*this.period;
		double tl_x = c_x - sampleRadius*period;
		double tl_y = c_y - sampleRadius*period;

		if (weights == null)
			return computeUnweighted(tl_x, tl_y, period);
		else
			return computeWeighted(tl_x, tl_y, period);
	}

	protected double computeUnweighted( double tl_x, double tl_y,
										double samplePeriod ) {
		// add 0.5 to c_x and c_y to have it round
		tl_x += 0.5;
		tl_y += 0.5;

		double Dx = 0, Dy = 0;
		int i = 0;
		for (int y = 0; y < sampleWidth; y++) {
			int pixelY = (int)(tl_y + y*samplePeriod);

			for (int x = 0; x < sampleWidth; x++, i++) {
				int pixelX = (int)(tl_x + x*samplePeriod);

				if (sampler.isInBounds(pixelX, pixelY)) {
					try {
						double val = sampler.compute(pixelX, pixelY);
						Dx += kerCosine.data[i]*val;
						Dy += kerSine.data[i]*val;
					} catch (RuntimeException e) {
						sampler.isInBounds(pixelX, pixelY);
						sampler.compute(pixelX, pixelY);
						throw e;
					}
				}
			}
		}

		return Math.atan2(Dy, Dx);
	}

	protected double computeWeighted( double tl_x, double tl_y,
									  double samplePeriod ) {
		// add 0.5 to c_x and c_y to have it round
		tl_x += 0.5;
		tl_y += 0.5;

		double Dx = 0, Dy = 0;
		int i = 0;
		for (int y = 0; y < sampleWidth; y++) {
			int pixelY = (int)(tl_y + y*samplePeriod);

			for (int x = 0; x < sampleWidth; x++, i++) {
				int pixelX = (int)(tl_x + x*samplePeriod);

				if (sampler.isInBounds(pixelX, pixelY)) {
					double val = sampler.compute(pixelX, pixelY);
					double w = weights.data[i];
					Dx += w*kerCosine.data[i]*val;
					Dy += w*kerSine.data[i]*val;
				}
			}
		}

		return Math.atan2(Dy, Dx);
	}

	@Override
	public RegionOrientation copy() {
		return new ImplOrientationImageAverageIntegral<>(
				objectRadiusToScale, sampleRadius, period, kernelWidth, weightSigma, getImageType());
	}
}