LucasKanadeFlow.evaluate

Lucas-Kanade optical flow algorithm implementation.

class LucasKanadeFlow

override

float[2][]

evaluate

(

inout Image f1

inout Image f2

in float[2][] points

in float[2][] searchRegions

float[2][] flow = null

bool usePrevious = false

)

in { assert (!f1.empty && !f2.empty && f1.size == f2.size && f1.channels == 1 && f1.depth == f2.depth); assert (points.length == searchRegions.length); if (usePrevious) { assert (flow !is null); assert (points.length == flow.length); } }

Parameters

f1: Type: Image
First frame image.
f2: Type: Image
Second frame image.
points: Type: float[2][]
points which are tracked.
searchRegions: Type: float[2][]
search region width and height for each point.
flow: Type: float[2][]
displacement values preallocated array.
usePrevious: Type: bool
if algorithm should continue iterating by using presented values in the flow array, set this to true.
See: dcv.features.corner

LucasKanadeFlow.evaluate

Parameters

Meta

Source