public class DenseOpticalFlowKlt<I extends boofcv.struct.image.ImageGray,D extends boofcv.struct.image.ImageGray>
extends java.lang.Object
KltTracker. A feature is computed from each pixel in the prev
image and it is tracked into the curr image. The flow assigned to a pixel is the template with the lowest error
which overlaps it. In other words, a pixel is assigned the flow with the lowest error with in 'radius' pixels
of it. A pixel is marked as invalid if all tracks around the pixel fail.| Constructor and Description |
|---|
DenseOpticalFlowKlt(PyramidKltTracker<I,D> tracker,
int numLayers,
int radius) |
| Modifier and Type | Method and Description |
|---|---|
protected void |
checkNeighbors(int cx,
int cy,
float score,
float flowX,
float flowY,
ImageFlow output)
Examines every pixel inside the region centered at (cx,cy) to see if their optical flow has a worse
score the one specified in 'flow'
|
void |
process(boofcv.struct.pyramid.ImagePyramid<I> prev,
D[] prevDerivX,
D[] prevDerivY,
boofcv.struct.pyramid.ImagePyramid<I> curr,
ImageFlow output) |
public DenseOpticalFlowKlt(PyramidKltTracker<I,D> tracker, int numLayers, int radius)
public void process(boofcv.struct.pyramid.ImagePyramid<I> prev, D[] prevDerivX, D[] prevDerivY, boofcv.struct.pyramid.ImagePyramid<I> curr, ImageFlow output)
protected void checkNeighbors(int cx,
int cy,
float score,
float flowX,
float flowY,
ImageFlow output)