FeatureTrackerConfig¶

This message is used to configure the FeatureTracker node. You can set the CornerDetector, FeatureMaintainer and MotionEstimator.

Reference¶

class depthai.FeatureTrackerConfig

FeatureTrackerConfig message. Carries config for feature tracking algorithm

class CornerDetector

Corner detector configuration structure.

class Thresholds

Threshold settings structure for corner detector.

property decreaseFactor: When detected number of features exceeds the maximum in a cell threshold is lowered by multiplying its value with this factor.

property increaseFactor: When detected number of features doesn’t exceed the maximum in a cell, threshold is increased by multiplying its value with this factor.

property initialValue: 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.

property max: Maximum limit for threshold. Applicable when automatic threshold update is enabled. 0 means auto. Empirical value.

property min: Minimum limit for threshold. Applicable when automatic threshold update is enabled. 0 means auto, 6000000 for HARRIS, 1200 for SHI_THOMASI. Empirical value.

class Type

Members:

HARRIS

SHI_THOMASI

property name

property cellGridDimension: 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.

property enableSobel: 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.

property enableSorting: Enable sorting detected features based on their score or not.

property numMaxFeatures: 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.

property numTargetFeatures: Target number of features to detect. Maximum number of features is determined at runtime based on algorithm type.

property thresholds: Threshold settings. These are advanced settings, suitable for debugging/special cases.

property type: Corner detector algorithm type.

class FeatureMaintainer

FeatureMaintainer configuration structure.

property enable: Enable feature maintaining or not.

property lostFeatureErrorThreshold: 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.

property minimumDistanceBetweenFeatures: 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.

property trackedFeatureThreshold: 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.

class MotionEstimator

Used for feature reidentification between current and previous features.

class OpticalFlow

Optical flow configuration structure.

property epsilon: 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.

property maxIterations: 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.

property pyramidLevels: 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.

property searchWindowHeight: 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

property searchWindowWidth: 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

class Type

Members:

LUCAS_KANADE_OPTICAL_FLOW

HW_MOTION_ESTIMATION

property name

property enable: Enable motion estimation or not.

property opticalFlow: Optical flow configuration. Takes effect only if MotionEstimator algorithm type set to LUCAS_KANADE_OPTICAL_FLOW.

property type: Motion estimator algorithm type.

get(self: depthai.FeatureTrackerConfig) → depthai.RawFeatureTrackerConfig

Retrieve configuration data for FeatureTracker.

Returns: config for feature tracking algorithm

getData(self: object) → numpy.ndarray[numpy.uint8]

Returns: Reference to internal buffer

getRaw(self: depthai.ADatatype) → depthai.RawBuffer

set(self: depthai.FeatureTrackerConfig, config: depthai.RawFeatureTrackerConfig) → None

Set explicit configuration.

Parameter config:: Explicit configuration

setCornerDetector(*args, **kwargs)

Overloaded function.

setCornerDetector(self: depthai.FeatureTrackerConfig, cornerDetector: depthai.RawFeatureTrackerConfig.CornerDetector.Type) -> None

Set corner detector algorithm type.

Parameter cornerDetector:: Corner detector type, HARRIS or SHI_THOMASI

setCornerDetector(self: depthai.FeatureTrackerConfig, config: depthai.RawFeatureTrackerConfig.CornerDetector) -> None

Set corner detector full configuration.

Parameter config:: Corner detector configuration

setData(*args, **kwargs)

Overloaded function.

setData(self: depthai.Buffer, arg0: List[int]) -> None

Parameter data:: Copies data to internal buffer

setData(self: depthai.Buffer, arg0: numpy.ndarray[numpy.uint8]) -> None

Parameter data:: Copies data to internal buffer

setFeatureMaintainer(*args, **kwargs)

Overloaded function.

setFeatureMaintainer(self: depthai.FeatureTrackerConfig, enable: bool) -> None

Enable or disable feature maintainer.

Parameter enable:

setFeatureMaintainer(self: depthai.FeatureTrackerConfig, config: depthai.RawFeatureTrackerConfig.FeatureMaintainer) -> None

Set feature maintainer full configuration.

Parameter config:: feature maintainer configuration

setHwMotionEstimation(self: depthai.FeatureTrackerConfig) → None: Set hardware accelerated motion estimation using block matching. Faster than optical flow (software implementation) but might not be as accurate.

setMotionEstimator(*args, **kwargs)

Overloaded function.

setMotionEstimator(self: depthai.FeatureTrackerConfig, enable: bool) -> None

Enable or disable motion estimator.

Parameter enable:

setMotionEstimator(self: depthai.FeatureTrackerConfig, config: depthai.RawFeatureTrackerConfig.MotionEstimator) -> None

Set motion estimator full configuration.

Parameter config:: Motion estimator configuration

setNumTargetFeatures(self: depthai.FeatureTrackerConfig, numTargetFeatures: int) → None

Set number of target features to detect.

Parameter numTargetFeatures:: Number of features

setOpticalFlow(*args, **kwargs)

Overloaded function.

setOpticalFlow(self: depthai.FeatureTrackerConfig) -> None

Set optical flow as motion estimation algorithm type.

setOpticalFlow(self: depthai.FeatureTrackerConfig, config: depthai.RawFeatureTrackerConfig.MotionEstimator.OpticalFlow) -> None

Set optical flow full configuration.

Parameter config:: Optical flow configuration

class dai::FeatureTrackerConfig : public dai::Buffer ¶

FeatureTrackerConfig message. Carries config for feature tracking algorithm

Public Types

using CornerDetector = RawFeatureTrackerConfig::CornerDetector ¶

using MotionEstimator = RawFeatureTrackerConfig::MotionEstimator ¶

using FeatureMaintainer = RawFeatureTrackerConfig::FeatureMaintainer ¶

Public Functions

FeatureTrackerConfig()¶: Construct FeatureTrackerConfig message.

FeatureTrackerConfig(std::shared_ptr<RawFeatureTrackerConfig> ptr)¶

~FeatureTrackerConfig() = default¶

void setCornerDetector(dai::FeatureTrackerConfig::CornerDetector::Type cornerDetector)¶

Set corner detector algorithm type.

Parameters

cornerDetector: Corner detector type, HARRIS or SHI_THOMASI

void setCornerDetector(dai::FeatureTrackerConfig::CornerDetector config)¶

Set corner detector full configuration.

Parameters

config: Corner detector configuration

void setOpticalFlow()¶: Set optical flow as motion estimation algorithm type.

void setOpticalFlow(dai::FeatureTrackerConfig::MotionEstimator::OpticalFlow config)¶

Set optical flow full configuration.

Parameters

config: Optical flow configuration

void setHwMotionEstimation()¶: Set hardware accelerated motion estimation using block matching. Faster than optical flow (software implementation) but might not be as accurate.

void setNumTargetFeatures(std::int32_t numTargetFeatures)¶

Set number of target features to detect.

Parameters