public AdaptiveSquareBinaryFilter(int radius, double bias, boolean down, ImageType<T> inputType) {
this.radius = radius;
this.bias = bias;
this.down = down;
this.inputType = inputType;
work1 = inputType.createImage(1, 1);
work2 = inputType.createImage(1, 1);
}
/**
* Creates a new image with all of its parameters initialized, including the data array.
*
* @param width Image's width.
* @param height Image's height.
* @param numBands Number of bands/colors.
*/
protected ImageInterleaved(int width, int height, int numBands) {
_setData(Array.newInstance(getPrimitiveDataType(), width * height * numBands));
this.startIndex = 0;
this.stride = width * numBands;
this.numBands = numBands;
this.width = width;
this.height = height;
this.imageType = (ImageType) ImageType.interleaved(numBands, getClass());
}
public SurfMultiSpectral_to_DescribeRegionPoint(
DescribePointSurfMultiSpectral<II> alg, Class<T> imageType, Class<II> integralType) {
this.alg = alg;
gray = GeneralizedImageOps.createSingleBand(imageType, 1, 1);
grayII = GeneralizedImageOps.createSingleBand(integralType, 1, 1);
bandII = new MultiSpectral<II>(integralType, 1, 1, alg.getNumBands());
this.imageType = ImageType.ms(alg.getNumBands(), imageType);
}
public static void main(String args[]) {
VisualizeCirculantTrackerApp app =
new VisualizeCirculantTrackerApp<ImageUInt8>(ImageUInt8.class);
// String fileName = "../data/applet/trackzing/track_peter.mjpeg";
String fileName = "../data/applet/tracking/day_follow_car.mjpeg";
SimpleImageSequence<ImageUInt8> sequence =
DefaultMediaManager.INSTANCE.openVideo(fileName, ImageType.single(ImageUInt8.class));
app.process(sequence);
}
public static void evaluate(String dataset) {
Class type = ImageFloat32.class;
DebugTldTrackerTldData generator = new DebugTldTrackerTldData(ImageType.single(type));
InterpolatePixelS interpolate = FactoryInterpolation.bilinearPixelS(type, BorderType.EXTENDED);
ImageGradient gradient = FactoryDerivative.sobel(type, type);
TldTracker tracker = new TldTracker(null, interpolate, gradient, type, type);
generator.evaluate(dataset, tracker);
}
protected MonocularPlaneVisualOdometry<ImageUInt8> createAlgorithm() {
PkltConfig config = new PkltConfig();
config.pyramidScaling = new int[] {1, 2, 4, 8};
config.templateRadius = 3;
ConfigGeneralDetector configDetector = new ConfigGeneralDetector(600, 3, 1);
PointTracker<ImageUInt8> tracker =
FactoryPointTracker.klt(config, configDetector, ImageUInt8.class, ImageSInt16.class);
return FactoryVisualOdometry.monoPlaneInfinity(
50, 2, 1.5, 300, tracker, ImageType.single(ImageUInt8.class));
}
private StitchingFromMotion2D<GrayU8, Affine2D_F64> createStabilization() {
ConfigGeneralDetector config = new ConfigGeneralDetector();
config.maxFeatures = 150;
config.threshold = 40;
config.radius = 3;
PointTracker<GrayU8> tracker =
FactoryPointTracker.klt(new int[] {1, 2, 4}, config, 3, GrayU8.class, GrayS16.class);
ImageMotion2D<GrayU8, Affine2D_F64> motion =
FactoryMotion2D.createMotion2D(
100, 1.5, 2, 40, 0.5, 0.6, false, tracker, new Affine2D_F64());
return FactoryMotion2D.createVideoStitch(0.2, motion, ImageType.single(GrayU8.class));
}
public void setDistorted(IntrinsicParameters param, DenseMatrix64F rect) {
if (rect == null) {
this.undoRadial =
LensDistortionOps.imageRemoveDistortion(
AdjustmentType.FULL_VIEW,
BorderType.VALUE,
param,
null,
ImageType.single(ImageFloat32.class));
this.remove_p_to_p =
LensDistortionOps.transform_F32(AdjustmentType.FULL_VIEW, param, null, false);
} else {
this.undoRadial =
RectifyImageOps.rectifyImage(param, rect, BorderType.VALUE, ImageFloat32.class);
this.remove_p_to_p = RectifyImageOps.transformPixelToRect_F32(param, rect);
}
}
public VisualizeAssociationMatchesApp(Class<T> imageType, Class<D> derivType) {
super(3);
this.imageType = imageType;
GeneralFeatureDetector<T, D> alg;
addAlgorithm(
0,
"Fast Hessian",
FactoryInterestPoint.fastHessian(new ConfigFastHessian(1, 2, 200, 1, 9, 4, 4)));
if (imageType == ImageFloat32.class)
addAlgorithm(
0, "SIFT", FactoryInterestPoint.siftDetector(null, new ConfigSiftDetector(2, 5, 200, 5)));
alg =
FactoryDetectPoint.createShiTomasi(new ConfigGeneralDetector(500, 2, 1), false, derivType);
addAlgorithm(0, "Shi-Tomasi", FactoryInterestPoint.wrapPoint(alg, 1, imageType, derivType));
addAlgorithm(1, "SURF-S", FactoryDescribeRegionPoint.surfStable(null, imageType));
addAlgorithm(
1,
"SURF-S Color",
FactoryDescribeRegionPoint.surfColorStable(null, ImageType.ms(3, imageType)));
if (imageType == ImageFloat32.class)
addAlgorithm(1, "SIFT", FactoryDescribeRegionPoint.sift(null, null));
addAlgorithm(1, "BRIEF", FactoryDescribeRegionPoint.brief(new ConfigBrief(true), imageType));
addAlgorithm(1, "BRIEFSO", FactoryDescribeRegionPoint.brief(new ConfigBrief(false), imageType));
addAlgorithm(1, "Pixel 11x11", FactoryDescribeRegionPoint.pixel(11, 11, imageType));
addAlgorithm(1, "NCC 11x11", FactoryDescribeRegionPoint.pixelNCC(11, 11, imageType));
addAlgorithm(2, "Greedy", false);
addAlgorithm(2, "Backwards", true);
// estimate orientation using this once since it is fast and accurate
Class integralType = GIntegralImageOps.getIntegralType(imageType);
OrientationIntegral orientationII = FactoryOrientationAlgs.sliding_ii(null, integralType);
orientation = FactoryOrientation.convertImage(orientationII, imageType);
imageLeft = new MultiSpectral<T>(imageType, 1, 1, 3);
imageRight = new MultiSpectral<T>(imageType, 1, 1, 3);
grayLeft = GeneralizedImageOps.createSingleBand(imageType, 1, 1);
grayRight = GeneralizedImageOps.createSingleBand(imageType, 1, 1);
setMainGUI(panel);
}
@Override
public ImageType<Output> getDerivativeType() {
return ImageType.single((Class) m.getParameterTypes()[2]);
}
protected ShowGradient() {
super(ImageType.single(GrayU8.class));
}
@Override
public ImageType<ImageUInt8> getOutputType() {
return ImageType.single(ImageUInt8.class);
}
@Override
public ImageType<T> getImageType() {
return ImageType.single(imageType);
}
@Override
public ImageType<ImageFloat32> getImageType() {
return ImageType.single(ImageFloat32.class);
}
public PointProcessing(StitchingFromMotion2D<GrayU8, Affine2D_F64> alg) {
super(ImageType.single(GrayU8.class));
this.alg = alg;
}
public TestImplImageDistort_I8() {
super(ImageType.single(GrayU8.class));
}
public static void main(String[] args) {
// Example with a moving camera. Highlights why motion estimation is sometimes required
String fileName = UtilIO.pathExample("tracking/chipmunk.mjpeg");
// Camera has a bit of jitter in it. Static kinda works but motion reduces false positives
// String fileName = UtilIO.pathExample("background/horse_jitter.mp4");
// Comment/Uncomment to switch input image type
ImageType imageType = ImageType.single(GrayF32.class);
// ImageType imageType = ImageType.il(3, InterleavedF32.class);
// ImageType imageType = ImageType.il(3, InterleavedU8.class);
// Configure the feature detector
ConfigGeneralDetector confDetector = new ConfigGeneralDetector();
confDetector.threshold = 10;
confDetector.maxFeatures = 300;
confDetector.radius = 6;
// Use a KLT tracker
PointTracker tracker =
FactoryPointTracker.klt(new int[] {1, 2, 4, 8}, confDetector, 3, GrayF32.class, null);
// This estimates the 2D image motion
ImageMotion2D<GrayF32, Homography2D_F64> motion2D =
FactoryMotion2D.createMotion2D(
500, 0.5, 3, 100, 0.6, 0.5, false, tracker, new Homography2D_F64());
ConfigBackgroundBasic configBasic = new ConfigBackgroundBasic(30, 0.005f);
// Configuration for Gaussian model. Note that the threshold changes depending on the number of
// image bands
// 12 = gray scale and 40 = color
ConfigBackgroundGaussian configGaussian = new ConfigBackgroundGaussian(12, 0.001f);
configGaussian.initialVariance = 64;
configGaussian.minimumDifference = 5;
// Comment/Uncomment to switch background mode
BackgroundModelMoving background =
FactoryBackgroundModel.movingBasic(
configBasic, new PointTransformHomography_F32(), imageType);
// FactoryBackgroundModel.movingGaussian(configGaussian, new PointTransformHomography_F32(),
// imageType);
MediaManager media = DefaultMediaManager.INSTANCE;
SimpleImageSequence video = media.openVideo(fileName, background.getImageType());
// media.openCamera(null,640,480,background.getImageType());
// ====== Initialize Images
// storage for segmented image. Background = 0, Foreground = 1
GrayU8 segmented = new GrayU8(video.getNextWidth(), video.getNextHeight());
// Grey scale image that's the input for motion estimation
GrayF32 grey = new GrayF32(segmented.width, segmented.height);
// coordinate frames
Homography2D_F32 firstToCurrent32 = new Homography2D_F32();
Homography2D_F32 homeToWorld = new Homography2D_F32();
homeToWorld.a13 = grey.width / 2;
homeToWorld.a23 = grey.height / 2;
// Create a background image twice the size of the input image. Tell it that the home is in the
// center
background.initialize(grey.width * 2, grey.height * 2, homeToWorld);
BufferedImage visualized =
new BufferedImage(segmented.width, segmented.height, BufferedImage.TYPE_INT_RGB);
ImageGridPanel gui = new ImageGridPanel(1, 2);
gui.setImages(visualized, visualized);
ShowImages.showWindow(gui, "Detections", true);
double fps = 0;
double alpha = 0.01; // smoothing factor for FPS
while (video.hasNext()) {
ImageBase input = video.next();
long before = System.nanoTime();
GConvertImage.convert(input, grey);
if (!motion2D.process(grey)) {
throw new RuntimeException("Should handle this scenario");
}
Homography2D_F64 firstToCurrent64 = motion2D.getFirstToCurrent();
UtilHomography.convert(firstToCurrent64, firstToCurrent32);
background.segment(firstToCurrent32, input, segmented);
background.updateBackground(firstToCurrent32, input);
long after = System.nanoTime();
fps = (1.0 - alpha) * fps + alpha * (1.0 / ((after - before) / 1e9));
VisualizeBinaryData.renderBinary(segmented, false, visualized);
gui.setImage(0, 0, (BufferedImage) video.getGuiImage());
gui.setImage(0, 1, visualized);
gui.repaint();
System.out.println("FPS = " + fps);
try {
Thread.sleep(5);
} catch (InterruptedException e) {
}
}
}
public DebugTldTrackerTldData(ImageType<T> type) {
input = type.createImage(1, 1);
}
