# FeatureTrackerConfig This message is used to configure the [FeatureTracker](https://docs.luxonis.com/software/depthai/examples/feature_tracker.md) node. You can set the CornerDetector, FeatureMaintainer and MotionEstimator. ## Reference ### depthai.FeatureTrackerConfig(depthai.Buffer) Kind: Class FeatureTrackerConfig message. Carries config for feature tracking algorithm #### depthai.FeatureTrackerConfig.CornerDetector Kind: Class Corner detector configuration structure. ##### depthai.FeatureTrackerConfig.CornerDetector.Thresholds Kind: Class Threshold settings structure for corner detector. ###### __init__(self: RawFeatureTrackerConfig.CornerDetector.Thresholds) Kind: Method ###### decreaseFactor Kind: Property When detected number of features exceeds the maximum in a cell threshold is lowered by multiplying its value with this factor. ###### decreaseFactor.setter(self, arg0: typing.SupportsFloat) Kind: Method ###### increaseFactor Kind: Property When detected number of features doesn't exceed the maximum in a cell, threshold is increased by multiplying its value with this factor. ###### increaseFactor.setter(self, arg0: typing.SupportsFloat) Kind: Method ###### initialValue Kind: Property Minimum strength of a feature which will be detected. 0 means automatic threshold update. Recommended so the tracker can adapt to different scenes/textures. Each cell has its own threshold. Empirical value. ###### initialValue.setter(self, arg0: typing.SupportsFloat) Kind: Method ###### max Kind: Property Maximum limit for threshold. Applicable when automatic threshold update is enabled. 0 means auto. Empirical value. ###### max.setter(self, arg0: typing.SupportsFloat) Kind: Method ###### min Kind: Property Minimum limit for threshold. Applicable when automatic threshold update is enabled. 0 means auto, 6000000 for HARRIS, 1200 for SHI_THOMASI. Empirical value. ###### min.setter(self, arg0: typing.SupportsFloat) Kind: Method ##### depthai.FeatureTrackerConfig.CornerDetector.Type Kind: Class Members: HARRIS SHI_THOMASI ###### HARRIS: typing.ClassVar[RawFeatureTrackerConfig.CornerDetector.Type] Kind: Class Variable ###### SHI_THOMASI: typing.ClassVar[RawFeatureTrackerConfig.CornerDetector.Type] Kind: Class Variable ###### __members__: typing.ClassVar[dict[str, RawFeatureTrackerConfig.CornerDetector.Type]] Kind: Class Variable ###### __eq__(self, other: typing.Any) -> bool: bool Kind: Method ###### __getstate__(self) -> int: int Kind: Method ###### __hash__(self) -> int: int Kind: Method ###### __index__(self: RawFeatureTrackerConfig.CornerDetector.Type) -> int: int Kind: Method ###### __init__(self: RawFeatureTrackerConfig.CornerDetector.Type, value: typing.SupportsInt) Kind: Method ###### __int__(self: RawFeatureTrackerConfig.CornerDetector.Type) -> int: int Kind: Method ###### __ne__(self, other: typing.Any) -> bool: bool Kind: Method ###### __repr__(self) -> str: str Kind: Method ###### __setstate__(self: RawFeatureTrackerConfig.CornerDetector.Type, state: typing.SupportsInt) Kind: Method ###### __str__(self) -> str: str Kind: Method ###### name Kind: Property ###### value Kind: Property ##### __init__(self: RawFeatureTrackerConfig.CornerDetector) Kind: Method ##### cellGridDimension Kind: Property Ensures distributed feature detection across the image. Image is divided into horizontal and vertical cells, each cell has a target feature count = numTargetFeatures / cellGridDimension. Each cell has its own feature threshold. A value of 4 means that the image is divided into 4x4 cells of equal width/height. Maximum 4, minimum 1. ##### cellGridDimension.setter(self, arg0: typing.SupportsInt) Kind: Method ##### enableSobel Kind: Property Enable 3x3 Sobel operator to smoothen the image whose gradient is to be computed. If disabled, a simple 1D row/column differentiator is used for gradient. ##### enableSobel.setter(self, arg0: bool) Kind: Method ##### enableSorting Kind: Property Enable sorting detected features based on their score or not. ##### enableSorting.setter(self, arg0: bool) Kind: Method ##### numMaxFeatures Kind: Property Hard limit for the maximum number of features that can be detected. 0 means auto, will be set to the maximum value based on memory constraints. ##### numMaxFeatures.setter(self, arg0: typing.SupportsInt) Kind: Method ##### numTargetFeatures Kind: Property Target number of features to detect. Maximum number of features is determined at runtime based on algorithm type. ##### numTargetFeatures.setter(self, arg0: typing.SupportsInt) Kind: Method ##### thresholds Kind: Property Threshold settings. These are advanced settings, suitable for debugging/special cases. ##### thresholds.setter(self, arg0: RawFeatureTrackerConfig.CornerDetector.Thresholds) Kind: Method ##### type Kind: Property Corner detector algorithm type. ##### type.setter(self, arg0: RawFeatureTrackerConfig.CornerDetector.Type) Kind: Method #### depthai.FeatureTrackerConfig.FeatureMaintainer Kind: Class FeatureMaintainer configuration structure. ##### __init__(self: RawFeatureTrackerConfig.FeatureMaintainer) Kind: Method ##### enable Kind: Property Enable feature maintaining or not. ##### enable.setter(self, arg0: bool) Kind: Method ##### lostFeatureErrorThreshold Kind: Property Optical flow measures the tracking error for every feature. If the point can’t be tracked or it’s out of the image it will set this error to a maximum value. This threshold defines the level where the tracking accuracy is considered too bad to keep the point. ##### lostFeatureErrorThreshold.setter(self, arg0: typing.SupportsFloat) Kind: Method ##### minimumDistanceBetweenFeatures Kind: Property Used to filter out detected feature points that are too close. Requires sorting enabled in detector. Unit of measurement is squared euclidean distance in pixels. ##### minimumDistanceBetweenFeatures.setter(self, arg0: typing.SupportsFloat) Kind: Method ##### trackedFeatureThreshold Kind: Property Once a feature was detected and we started tracking it, we need to update its Harris score on each image. This is needed because a feature point can disappear, or it can become too weak to be tracked. This threshold defines the point where such a feature must be dropped. As the goal of the algorithm is to provide longer tracks, we try to add strong points and track them until they are absolutely untrackable. This is why, this value is usually smaller than the detection threshold. ##### trackedFeatureThreshold.setter(self, arg0: typing.SupportsFloat) Kind: Method #### depthai.FeatureTrackerConfig.MotionEstimator Kind: Class Used for feature reidentification between current and previous features. ##### depthai.FeatureTrackerConfig.MotionEstimator.OpticalFlow Kind: Class Optical flow configuration structure. ###### __init__(self: RawFeatureTrackerConfig.MotionEstimator.OpticalFlow) Kind: Method ###### epsilon Kind: Property Feature tracking termination criteria. Optical flow will refine the feature position on each pyramid level until the displacement between two refinements is smaller than this value. Decreasing this number increases runtime. ###### epsilon.setter(self, arg0: typing.SupportsFloat) Kind: Method ###### maxIterations Kind: Property Feature tracking termination criteria. Optical flow will refine the feature position maximum this many times on each pyramid level. If the Epsilon criteria described in the previous chapter is not met after this number of iterations, the algorithm will continue with the current calculated value. Increasing this number increases runtime. ###### maxIterations.setter(self, arg0: typing.SupportsInt) Kind: Method ###### pyramidLevels Kind: Property Number of pyramid levels, only for optical flow. AUTO means it's decided based on input resolution: 3 if image width <= 640, else 4. Valid values are either 3/4 for VGA, 4 for 720p and above. ###### pyramidLevels.setter(self, arg0: typing.SupportsInt) Kind: Method ###### searchWindowHeight Kind: Property Image patch height used to track features. Must be an odd number, maximum 9. N means the algorithm will be able to track motion at most (N-1)/2 pixels in a direction per pyramid level. Increasing this number increases runtime ###### searchWindowHeight.setter(self, arg0: typing.SupportsInt) Kind: Method ###### searchWindowWidth Kind: Property Image patch width used to track features. Must be an odd number, maximum 9. N means the algorithm will be able to track motion at most (N-1)/2 pixels in a direction per pyramid level. Increasing this number increases runtime ###### searchWindowWidth.setter(self, arg0: typing.SupportsInt) Kind: Method ##### depthai.FeatureTrackerConfig.MotionEstimator.Type Kind: Class Members: LUCAS_KANADE_OPTICAL_FLOW HW_MOTION_ESTIMATION ###### HW_MOTION_ESTIMATION: typing.ClassVar[RawFeatureTrackerConfig.MotionEstimator.Type] Kind: Class Variable ###### LUCAS_KANADE_OPTICAL_FLOW: typing.ClassVar[RawFeatureTrackerConfig.MotionEstimator.Type] Kind: Class Variable ###### __members__: typing.ClassVar[dict[str, RawFeatureTrackerConfig.MotionEstimator.Type]] Kind: Class Variable ###### __eq__(self, other: typing.Any) -> bool: bool Kind: Method ###### __getstate__(self) -> int: int Kind: Method ###### __hash__(self) -> int: int Kind: Method ###### __index__(self: RawFeatureTrackerConfig.MotionEstimator.Type) -> int: int Kind: Method ###### __init__(self: RawFeatureTrackerConfig.MotionEstimator.Type, value: typing.SupportsInt) Kind: Method ###### __int__(self: RawFeatureTrackerConfig.MotionEstimator.Type) -> int: int Kind: Method ###### __ne__(self, other: typing.Any) -> bool: bool Kind: Method ###### __repr__(self) -> str: str Kind: Method ###### __setstate__(self: RawFeatureTrackerConfig.MotionEstimator.Type, state: typing.SupportsInt) Kind: Method ###### __str__(self) -> str: str Kind: Method ###### name Kind: Property ###### value Kind: Property ##### __init__(self: RawFeatureTrackerConfig.MotionEstimator) Kind: Method ##### enable Kind: Property Enable motion estimation or not. ##### enable.setter(self, arg0: bool) Kind: Method ##### opticalFlow Kind: Property Optical flow configuration. Takes effect only if MotionEstimator algorithm type set to LUCAS_KANADE_OPTICAL_FLOW. ##### opticalFlow.setter(self, arg0: RawFeatureTrackerConfig.MotionEstimator.OpticalFlow) Kind: Method ##### type Kind: Property Motion estimator algorithm type. ##### type.setter(self, arg0: RawFeatureTrackerConfig.MotionEstimator.Type) Kind: Method #### __init__() Kind: Method #### get(self) -> RawFeatureTrackerConfig: RawFeatureTrackerConfig Kind: Method Retrieve configuration data for FeatureTracker. Returns: config for feature tracking algorithm #### set(self, config: RawFeatureTrackerConfig) -> FeatureTrackerConfig: FeatureTrackerConfig Kind: Method Set explicit configuration. Parameter ``config``: Explicit configuration #### setCornerDetector() Kind: Method #### setFeatureMaintainer() Kind: Method #### setHwMotionEstimation(self) -> FeatureTrackerConfig: FeatureTrackerConfig Kind: Method Set hardware accelerated motion estimation using block matching. Faster than optical flow (software implementation) but might not be as accurate. #### setMotionEstimator() Kind: Method #### setNumTargetFeatures(self, numTargetFeatures: typing.SupportsInt) -> FeatureTrackerConfig: FeatureTrackerConfig Kind: Method Set number of target features to detect. Parameter ``numTargetFeatures``: Number of features #### setOpticalFlow() Kind: Method ### Need assistance? Head over to [Discussion Forum](https://discuss.luxonis.com/) for technical support or any other questions you might have.