TrackerMIL

Motion and trackingclassOpenCV 5.0.0

import { TrackerMIL } from '@banou/opencv-wasm'

Use after await initOpenCV(). See the initialization and named imports guide.

Explore visually: MIL tracking →

Learn an online appearance classifier from bags of nearby positive and negative samples.

VISUAL WALKTHROUGHMotion

REFERENCE INPUTTarget moves in the next frame

The previous position is a starting estimate, not the new target location.

→

RESULTUpdate the target box

Choose the best-scoring patch in the next search area.

Illustrative example. The stages explain the method; they are not a live OpenCV execution.

Native object: release it with using or delete(). Factories can return null; check before calling methods. Inherits Tracker.

The MIL algorithm trains a classifier in an online manner to separate the object from the background.

Multiple Instance Learning avoids the drift problem for a robust tracking. The implementation is based on [MIL] .

Original code can be found here http://vision.ucsd.edu/~bbabenko/project_miltrack.shtml

Upstream declaration ↗

Constructors and members

static create

Create MIL tracker instance

create(parameters: TrackerMIL_Params): TrackerMIL | null;

2 available overloads

create(): TrackerMIL | null;

create(parameters: TrackerMIL_Params): TrackerMIL | null;

parameters

MIL parameters TrackerMIL::Params

Returns

The TrackerMIL | null result.

Upstream declaration ↗

clone

Create another handle to the same native object. This retains the object without copying its pixels or algorithm state; dispose both handles separately.

clone(): this;

Returns

The this result.

These signatures describe this package. Upstream documentation can mention optional backends that are absent from this build. Check runtime compatibility before choosing a backend or file format.