This class implement the feature tracker module based on Lucas-Kanade optical flow. The tracker tracks feature temporally in subsequent images and produces a set of matches. More...

#include <tracker.hpp>

Public Types
using	Keypoints = std::vector<cv::KeyPoint>
	A vector of features keypoints.

using	TimedFeatures = std::pair<fp, Features>
	Set of features associated with a time.

Public Member Functions
	Tracker (const TrackerOptions &opts, const Vector4 &intrinsics)
	Tracker constructor.

void	processCamera (Camera &cam)
	This method process the input camera measurement. If first pre-process the camera image, and then it tracks features.

const TimedFeatures &	currentFeatures () const
	Get the current detected/tracked features.

const PinholeCameraUniquePtr &	cam () const
	Get the camera pointer.

Detailed Description

This class implement the feature tracker module based on Lucas-Kanade optical flow. The tracker tracks feature temporally in subsequent images and produces a set of matches.

Constructor & Destructor Documentation

◆ Tracker()

msceqf::Tracker::Tracker	(	const TrackerOptions &	opts,
		const Vector4 &	intrinsics )

Tracker constructor.

Parameters

opts	Tracker options
intrinsics	Camera intrinsics as R4 vector (fx, fy, cx, cy)

Member Function Documentation

◆ cam()

const PinholeCameraUniquePtr & msceqf::Tracker::cam ( ) const

Get the camera pointer.

Returns: Pointer to the camera object

◆ currentFeatures()

const TimedFeatures & msceqf::Tracker::currentFeatures ( ) const

Get the current detected/tracked features.

Returns: Current detected/tracked features

◆ processCamera()

void msceqf::Tracker::processCamera ( Camera & cam )

This method process the input camera measurement. If first pre-process the camera image, and then it tracks features.