boofcv.alg.feature.detect.extract.NonMaxBlockSearchRelaxed 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
Show all versions of boofcv-feature Show documentation
BoofCV is an open source Java library for real-time computer vision and robotics applications.
/*
* Copyright (c) 2011-2019, 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.detect.extract;
import boofcv.struct.QueueCorner;
import boofcv.struct.image.GrayF32;
import georegression.struct.point.Point2D_I32;
/**
*
* Implementation of {@link NonMaxBlock} which implements a relaxed maximum rule.
*
*
* @author Peter Abeles
*/
public abstract class NonMaxBlockSearchRelaxed implements NonMaxBlock.Search {
// storage for local maximums
Point2D_I32 foundMax[];
Point2D_I32 foundMin[];
// threshold for intensity values when detecting minimums and maximums
float thresholdMin;
float thresholdMax;
int radius;
private QueueCorner localMin,localMax;
GrayF32 img;
@Override
public void initialize(NonMaxBlock.Configuration configuration, GrayF32 image,
QueueCorner localMin, QueueCorner localMax) {
this.thresholdMin = configuration.thresholdMin;
this.thresholdMax = configuration.thresholdMax;
this.radius = configuration.radius;
this.img = image;
this.localMin = localMin;
this.localMax = localMax;
int w = 2* radius +1;
// only declare this work space if needed
if( foundMax == null || foundMax.length != w*w ) {
foundMax = new Point2D_I32[w * w];
for (int i = 0; i < foundMax.length; i++)
foundMax[i] = new Point2D_I32();
foundMin = new Point2D_I32[w * w];
for (int i = 0; i < foundMin.length; i++)
foundMin[i] = new Point2D_I32();
}
}
public static class Max extends NonMaxBlockSearchRelaxed {
@Override
public void searchBlock( int x0 , int y0 , int x1 , int y1 ) {
int numPeaks = 0;
float peakVal = thresholdMax;
for( int y = y0; y < y1; y++ ) {
int index = img.startIndex + y*img.stride+x0;
for( int x = x0; x < x1; x++ ) {
float v = img.data[index++];
if( v > peakVal ) {
peakVal = v;
foundMax[0].set(x, y);
numPeaks = 1;
} else if( v == peakVal ) {
foundMax[numPeaks++].set(x, y);
}
}
}
if( numPeaks > 0 && peakVal != Float.MAX_VALUE ) {
for( int i = 0; i < numPeaks; i++ ) {
Point2D_I32 p = foundMax[i];
checkLocalMax(p.x,p.y,peakVal,img);
}
}
}
@Override
public boolean isDetectMinimums() {
return false;
}
@Override
public boolean isDetectMaximums() {
return true;
}
@Override
public NonMaxBlock.Search newInstance() {
return new Max();
}
}
public static class Min extends NonMaxBlockSearchRelaxed {
@Override
public void searchBlock( int x0 , int y0 , int x1 , int y1 ) {
int numPeaks = 0;
float peakVal = thresholdMin;
for( int y = y0; y < y1; y++ ) {
int index = img.startIndex + y*img.stride+x0;
for( int x = x0; x < x1; x++ ) {
float v = img.data[index++];
if( v < peakVal ) {
peakVal = v;
foundMin[0].set(x, y);
numPeaks = 1;
} else if( v == peakVal ) {
foundMin[numPeaks++].set(x, y);
}
}
}
if( numPeaks > 0 && peakVal != -Float.MAX_VALUE ) {
for( int i = 0; i < numPeaks; i++ ) {
Point2D_I32 p = foundMin[i];
checkLocalMin(p.x, p.y, peakVal, img);
}
}
}
@Override
public boolean isDetectMinimums() {
return true;
}
@Override
public boolean isDetectMaximums() {
return false;
}
@Override
public NonMaxBlock.Search newInstance() {
return new Min();
}
}
public static class MinMax extends NonMaxBlockSearchRelaxed {
@Override
public void searchBlock( int x0 , int y0 , int x1 , int y1 ) {
int numMinPeaks = 0;
float peakMinVal = thresholdMin;
int numMaxPeaks = 0;
float peakMaxVal = thresholdMax;
for( int y = y0; y < y1; y++ ) {
int index = img.startIndex + y*img.stride+x0;
for( int x = x0; x < x1; x++ ) {
float v = img.data[index++];
if( v < peakMinVal ) {
peakMinVal = v;
foundMin[0].set(x, y);
numMinPeaks = 1;
} else if( v == peakMinVal ) {
foundMin[numMinPeaks++].set(x, y);
}
if( v > peakMaxVal ) {
peakMaxVal = v;
foundMax[0].set(x, y);
numMaxPeaks = 1;
} else if( v == peakMaxVal ) {
foundMax[numMaxPeaks++].set(x, y);
}
}
}
if( numMinPeaks > 0 && peakMinVal != -Float.MAX_VALUE ) {
for( int i = 0; i < numMinPeaks; i++ ) {
Point2D_I32 p = foundMin[i];
checkLocalMin(p.x,p.y,peakMinVal,img);
}
}
if( numMaxPeaks > 0 && peakMaxVal != Float.MAX_VALUE ) {
for( int i = 0; i < numMaxPeaks; i++ ) {
Point2D_I32 p = foundMax[i];
checkLocalMax(p.x,p.y,peakMaxVal,img);
}
}
}
@Override
public boolean isDetectMinimums() {
return true;
}
@Override
public boolean isDetectMaximums() {
return true;
}
@Override
public NonMaxBlock.Search newInstance() {
return new MinMax();
}
}
protected void checkLocalMax( int x_c , int y_c , float peakVal , GrayF32 img ) {
int x0 = x_c-radius;
int x1 = x_c+radius;
int y0 = y_c-radius;
int y1 = y_c+radius;
if (x0 < 0) x0 = 0;
if (y0 < 0) y0 = 0;
if (x1 >= img.width) x1 = img.width - 1;
if (y1 >= img.height) y1 = img.height - 1;
for( int y = y0; y <= y1; y++ ) {
int index = img.startIndex + y*img.stride+x0;
for( int x = x0; x <= x1; x++ ) {
float v = img.data[index++];
if( v > peakVal ) {
// not a local maximum
return;
}
}
}
localMax.add(x_c,y_c);
}
protected void checkLocalMin( int x_c , int y_c , float peakVal , GrayF32 img ) {
int x0 = x_c-radius;
int x1 = x_c+radius;
int y0 = y_c-radius;
int y1 = y_c+radius;
if (x0 < 0) x0 = 0;
if (y0 < 0) y0 = 0;
if (x1 >= img.width) x1 = img.width - 1;
if (y1 >= img.height) y1 = img.height - 1;
for( int y = y0; y <= y1; y++ ) {
int index = img.startIndex + y*img.stride+x0;
for( int x = x0; x <= x1; x++ ) {
float v = img.data[index++];
if( v < peakVal ) {
// not a local minimum
return;
}
}
}
localMin.add(x_c,y_c);
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy