Saltar al contenido

Referencia para ultralytics/trackers/utils/kalman_filter.py

Nota

Este archivo est√° disponible en https://github.com/ultralytics/ ultralytics/blob/main/ ultralytics/trackers/utils/kalman_filter .py. Si detectas alg√ļn problema, por favor, ayuda a solucionarlo contribuyendo con una Pull Request ūüõ†ÔłŹ. ¬°Gracias ūüôŹ!



ultralytics.trackers.utils.kalman_filter.KalmanFilterXYAH

Para bytetrack. Un sencillo filtro de Kalman para el seguimiento de cajas delimitadoras en el espacio de la imagen.

El espacio de estado de 8 dimensiones (x, y, a, h, vx, vy, va, vh) contiene la posición central de la caja delimitadora (x, y), el aspecto a, altura h, y sus respectivas velocidades.

El movimiento del objeto sigue un modelo de velocidad constante. La posición de la caja delimitadora (x, y, a, h) se toma como observación directa del espacio de estado (modelo de observación lineal). del espacio de estado (modelo de observación lineal).

Código fuente en ultralytics/trackers/utils/kalman_filter.py
class KalmanFilterXYAH:
    """
    For bytetrack. A simple Kalman filter for tracking bounding boxes in image space.

    The 8-dimensional state space (x, y, a, h, vx, vy, va, vh) contains the bounding box center position (x, y), aspect
    ratio a, height h, and their respective velocities.

    Object motion follows a constant velocity model. The bounding box location (x, y, a, h) is taken as direct
    observation of the state space (linear observation model).
    """

    def __init__(self):
        """Initialize Kalman filter model matrices with motion and observation uncertainty weights."""
        ndim, dt = 4, 1.0

        # Create Kalman filter model matrices
        self._motion_mat = np.eye(2 * ndim, 2 * ndim)
        for i in range(ndim):
            self._motion_mat[i, ndim + i] = dt
        self._update_mat = np.eye(ndim, 2 * ndim)

        # Motion and observation uncertainty are chosen relative to the current state estimate. These weights control
        # the amount of uncertainty in the model.
        self._std_weight_position = 1.0 / 20
        self._std_weight_velocity = 1.0 / 160

    def initiate(self, measurement: np.ndarray) -> tuple:
        """
        Create track from unassociated measurement.

        Args:
            measurement (ndarray): Bounding box coordinates (x, y, a, h) with center position (x, y), aspect ratio a,
                and height h.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector (8 dimensional) and covariance matrix (8x8 dimensional)
                of the new track. Unobserved velocities are initialized to 0 mean.
        """
        mean_pos = measurement
        mean_vel = np.zeros_like(mean_pos)
        mean = np.r_[mean_pos, mean_vel]

        std = [
            2 * self._std_weight_position * measurement[3],
            2 * self._std_weight_position * measurement[3],
            1e-2,
            2 * self._std_weight_position * measurement[3],
            10 * self._std_weight_velocity * measurement[3],
            10 * self._std_weight_velocity * measurement[3],
            1e-5,
            10 * self._std_weight_velocity * measurement[3],
        ]
        covariance = np.diag(np.square(std))
        return mean, covariance

    def predict(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
        """
        Run Kalman filter prediction step.

        Args:
            mean (ndarray): The 8 dimensional mean vector of the object state at the previous time step.
            covariance (ndarray): The 8x8 dimensional covariance matrix of the object state at the previous time step.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
                velocities are initialized to 0 mean.
        """
        std_pos = [
            self._std_weight_position * mean[3],
            self._std_weight_position * mean[3],
            1e-2,
            self._std_weight_position * mean[3],
        ]
        std_vel = [
            self._std_weight_velocity * mean[3],
            self._std_weight_velocity * mean[3],
            1e-5,
            self._std_weight_velocity * mean[3],
        ]
        motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))

        mean = np.dot(mean, self._motion_mat.T)
        covariance = np.linalg.multi_dot((self._motion_mat, covariance, self._motion_mat.T)) + motion_cov

        return mean, covariance

    def project(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
        """
        Project state distribution to measurement space.

        Args:
            mean (ndarray): The state's mean vector (8 dimensional array).
            covariance (ndarray): The state's covariance matrix (8x8 dimensional).

        Returns:
            (tuple[ndarray, ndarray]): Returns the projected mean and covariance matrix of the given state estimate.
        """
        std = [
            self._std_weight_position * mean[3],
            self._std_weight_position * mean[3],
            1e-1,
            self._std_weight_position * mean[3],
        ]
        innovation_cov = np.diag(np.square(std))

        mean = np.dot(self._update_mat, mean)
        covariance = np.linalg.multi_dot((self._update_mat, covariance, self._update_mat.T))
        return mean, covariance + innovation_cov

    def multi_predict(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
        """
        Run Kalman filter prediction step (Vectorized version).

        Args:
            mean (ndarray): The Nx8 dimensional mean matrix of the object states at the previous time step.
            covariance (ndarray): The Nx8x8 covariance matrix of the object states at the previous time step.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
                velocities are initialized to 0 mean.
        """
        std_pos = [
            self._std_weight_position * mean[:, 3],
            self._std_weight_position * mean[:, 3],
            1e-2 * np.ones_like(mean[:, 3]),
            self._std_weight_position * mean[:, 3],
        ]
        std_vel = [
            self._std_weight_velocity * mean[:, 3],
            self._std_weight_velocity * mean[:, 3],
            1e-5 * np.ones_like(mean[:, 3]),
            self._std_weight_velocity * mean[:, 3],
        ]
        sqr = np.square(np.r_[std_pos, std_vel]).T

        motion_cov = [np.diag(sqr[i]) for i in range(len(mean))]
        motion_cov = np.asarray(motion_cov)

        mean = np.dot(mean, self._motion_mat.T)
        left = np.dot(self._motion_mat, covariance).transpose((1, 0, 2))
        covariance = np.dot(left, self._motion_mat.T) + motion_cov

        return mean, covariance

    def update(self, mean: np.ndarray, covariance: np.ndarray, measurement: np.ndarray) -> tuple:
        """
        Run Kalman filter correction step.

        Args:
            mean (ndarray): The predicted state's mean vector (8 dimensional).
            covariance (ndarray): The state's covariance matrix (8x8 dimensional).
            measurement (ndarray): The 4 dimensional measurement vector (x, y, a, h), where (x, y) is the center
                position, a the aspect ratio, and h the height of the bounding box.

        Returns:
            (tuple[ndarray, ndarray]): Returns the measurement-corrected state distribution.
        """
        projected_mean, projected_cov = self.project(mean, covariance)

        chol_factor, lower = scipy.linalg.cho_factor(projected_cov, lower=True, check_finite=False)
        kalman_gain = scipy.linalg.cho_solve(
            (chol_factor, lower), np.dot(covariance, self._update_mat.T).T, check_finite=False
        ).T
        innovation = measurement - projected_mean

        new_mean = mean + np.dot(innovation, kalman_gain.T)
        new_covariance = covariance - np.linalg.multi_dot((kalman_gain, projected_cov, kalman_gain.T))
        return new_mean, new_covariance

    def gating_distance(
        self,
        mean: np.ndarray,
        covariance: np.ndarray,
        measurements: np.ndarray,
        only_position: bool = False,
        metric: str = "maha",
    ) -> np.ndarray:
        """
        Compute gating distance between state distribution and measurements. A suitable distance threshold can be
        obtained from `chi2inv95`. If `only_position` is False, the chi-square distribution has 4 degrees of freedom,
        otherwise 2.

        Args:
            mean (ndarray): Mean vector over the state distribution (8 dimensional).
            covariance (ndarray): Covariance of the state distribution (8x8 dimensional).
            measurements (ndarray): An Nx4 matrix of N measurements, each in format (x, y, a, h) where (x, y)
                is the bounding box center position, a the aspect ratio, and h the height.
            only_position (bool, optional): If True, distance computation is done with respect to the bounding box
                center position only. Defaults to False.
            metric (str, optional): The metric to use for calculating the distance. Options are 'gaussian' for the
                squared Euclidean distance and 'maha' for the squared Mahalanobis distance. Defaults to 'maha'.

        Returns:
            (np.ndarray): Returns an array of length N, where the i-th element contains the squared distance between
                (mean, covariance) and `measurements[i]`.
        """
        mean, covariance = self.project(mean, covariance)
        if only_position:
            mean, covariance = mean[:2], covariance[:2, :2]
            measurements = measurements[:, :2]

        d = measurements - mean
        if metric == "gaussian":
            return np.sum(d * d, axis=1)
        elif metric == "maha":
            cholesky_factor = np.linalg.cholesky(covariance)
            z = scipy.linalg.solve_triangular(cholesky_factor, d.T, lower=True, check_finite=False, overwrite_b=True)
            return np.sum(z * z, axis=0)  # square maha
        else:
            raise ValueError("Invalid distance metric")

__init__()

Inicializa las matrices del modelo del filtro de Kalman con los pesos de la incertidumbre de movimiento y observación.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def __init__(self):
    """Initialize Kalman filter model matrices with motion and observation uncertainty weights."""
    ndim, dt = 4, 1.0

    # Create Kalman filter model matrices
    self._motion_mat = np.eye(2 * ndim, 2 * ndim)
    for i in range(ndim):
        self._motion_mat[i, ndim + i] = dt
    self._update_mat = np.eye(ndim, 2 * ndim)

    # Motion and observation uncertainty are chosen relative to the current state estimate. These weights control
    # the amount of uncertainty in the model.
    self._std_weight_position = 1.0 / 20
    self._std_weight_velocity = 1.0 / 160

gating_distance(mean, covariance, measurements, only_position=False, metric='maha')

Calcula la distancia entre la distribución de estados y las mediciones. Se puede obtener un umbral de distancia adecuado obtenerse a partir de chi2inv95. Si only_position es Falso, la distribución chi-cuadrado tiene 4 grados de libertad, en caso contrario, 2.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

Vector medio sobre la distribución de estados (8 dimensiones).

necesario
covariance ndarray

Covarianza de la distribución de estados (8x8 dimensiones).

necesario
measurements ndarray

Una matriz Nx4 de N medidas, cada una en formato (x, y, a, h) donde (x, y) es la posición central del cuadro delimitador, a la relación de aspecto y h la altura.

necesario
only_position bool

Si es Verdadero, el cálculo de la distancia se realiza con respecto a la posición central del cuadro delimitador sólo la posición central. Por defecto es Falso.

False
metric str

La métrica a utilizar para calcular la distancia. Las opciones son 'gaussian' para la distancia euclídea al cuadrado y "maha" para la distancia Mahalanobis al cuadrado. Por defecto es "maha".

'maha'

Devuelve:

Tipo Descripción
ndarray

Devuelve una matriz de longitud N, donde el i-ésimo elemento contiene la distancia al cuadrado entre (media, covarianza) y measurements[i].

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def gating_distance(
    self,
    mean: np.ndarray,
    covariance: np.ndarray,
    measurements: np.ndarray,
    only_position: bool = False,
    metric: str = "maha",
) -> np.ndarray:
    """
    Compute gating distance between state distribution and measurements. A suitable distance threshold can be
    obtained from `chi2inv95`. If `only_position` is False, the chi-square distribution has 4 degrees of freedom,
    otherwise 2.

    Args:
        mean (ndarray): Mean vector over the state distribution (8 dimensional).
        covariance (ndarray): Covariance of the state distribution (8x8 dimensional).
        measurements (ndarray): An Nx4 matrix of N measurements, each in format (x, y, a, h) where (x, y)
            is the bounding box center position, a the aspect ratio, and h the height.
        only_position (bool, optional): If True, distance computation is done with respect to the bounding box
            center position only. Defaults to False.
        metric (str, optional): The metric to use for calculating the distance. Options are 'gaussian' for the
            squared Euclidean distance and 'maha' for the squared Mahalanobis distance. Defaults to 'maha'.

    Returns:
        (np.ndarray): Returns an array of length N, where the i-th element contains the squared distance between
            (mean, covariance) and `measurements[i]`.
    """
    mean, covariance = self.project(mean, covariance)
    if only_position:
        mean, covariance = mean[:2], covariance[:2, :2]
        measurements = measurements[:, :2]

    d = measurements - mean
    if metric == "gaussian":
        return np.sum(d * d, axis=1)
    elif metric == "maha":
        cholesky_factor = np.linalg.cholesky(covariance)
        z = scipy.linalg.solve_triangular(cholesky_factor, d.T, lower=True, check_finite=False, overwrite_b=True)
        return np.sum(z * z, axis=0)  # square maha
    else:
        raise ValueError("Invalid distance metric")

initiate(measurement)

Crea una pista a partir de una medición no asociada.

Par√°metros:

Nombre Tipo Descripción Por defecto
measurement ndarray

Coordenadas de la caja delimitadora (x, y, a, h) con posición central (x, y), relación de aspecto a, y altura h.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media (8 dimensiones) y la matriz de covarianza (8x8 dimensiones) de la nueva pista. Las velocidades no observadas se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def initiate(self, measurement: np.ndarray) -> tuple:
    """
    Create track from unassociated measurement.

    Args:
        measurement (ndarray): Bounding box coordinates (x, y, a, h) with center position (x, y), aspect ratio a,
            and height h.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector (8 dimensional) and covariance matrix (8x8 dimensional)
            of the new track. Unobserved velocities are initialized to 0 mean.
    """
    mean_pos = measurement
    mean_vel = np.zeros_like(mean_pos)
    mean = np.r_[mean_pos, mean_vel]

    std = [
        2 * self._std_weight_position * measurement[3],
        2 * self._std_weight_position * measurement[3],
        1e-2,
        2 * self._std_weight_position * measurement[3],
        10 * self._std_weight_velocity * measurement[3],
        10 * self._std_weight_velocity * measurement[3],
        1e-5,
        10 * self._std_weight_velocity * measurement[3],
    ]
    covariance = np.diag(np.square(std))
    return mean, covariance

multi_predict(mean, covariance)

Ejecuta el paso de predicción del filtro Kalman (versión vectorizada).

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

La matriz media de Nx8 dimensiones de los estados del objeto en el paso de tiempo anterior.

necesario
covariance ndarray

La matriz de covarianza Nx8x8 de los estados del objeto en el paso de tiempo anterior.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media y la matriz de covarianza del estado predicho. Las velocidades no observadas de se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def multi_predict(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
    """
    Run Kalman filter prediction step (Vectorized version).

    Args:
        mean (ndarray): The Nx8 dimensional mean matrix of the object states at the previous time step.
        covariance (ndarray): The Nx8x8 covariance matrix of the object states at the previous time step.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
            velocities are initialized to 0 mean.
    """
    std_pos = [
        self._std_weight_position * mean[:, 3],
        self._std_weight_position * mean[:, 3],
        1e-2 * np.ones_like(mean[:, 3]),
        self._std_weight_position * mean[:, 3],
    ]
    std_vel = [
        self._std_weight_velocity * mean[:, 3],
        self._std_weight_velocity * mean[:, 3],
        1e-5 * np.ones_like(mean[:, 3]),
        self._std_weight_velocity * mean[:, 3],
    ]
    sqr = np.square(np.r_[std_pos, std_vel]).T

    motion_cov = [np.diag(sqr[i]) for i in range(len(mean))]
    motion_cov = np.asarray(motion_cov)

    mean = np.dot(mean, self._motion_mat.T)
    left = np.dot(self._motion_mat, covariance).transpose((1, 0, 2))
    covariance = np.dot(left, self._motion_mat.T) + motion_cov

    return mean, covariance

predict(mean, covariance)

Ejecuta el paso de predicción del filtro Kalman.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio de 8 dimensiones del estado del objeto en el paso de tiempo anterior.

necesario
covariance ndarray

La matriz de covarianza de 8x8 dimensiones del estado del objeto en el paso de tiempo anterior.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media y la matriz de covarianza del estado predicho. Las velocidades no observadas de se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def predict(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
    """
    Run Kalman filter prediction step.

    Args:
        mean (ndarray): The 8 dimensional mean vector of the object state at the previous time step.
        covariance (ndarray): The 8x8 dimensional covariance matrix of the object state at the previous time step.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
            velocities are initialized to 0 mean.
    """
    std_pos = [
        self._std_weight_position * mean[3],
        self._std_weight_position * mean[3],
        1e-2,
        self._std_weight_position * mean[3],
    ]
    std_vel = [
        self._std_weight_velocity * mean[3],
        self._std_weight_velocity * mean[3],
        1e-5,
        self._std_weight_velocity * mean[3],
    ]
    motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))

    mean = np.dot(mean, self._motion_mat.T)
    covariance = np.linalg.multi_dot((self._motion_mat, covariance, self._motion_mat.T)) + motion_cov

    return mean, covariance

project(mean, covariance)

Proyecta la distribución de estados al espacio de medida.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio del estado (matriz de 8 dimensiones).

necesario
covariance ndarray

La matriz de covarianza del estado (8x8 dimensiones).

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve la media proyectada y la matriz de covarianza de la estimación de estado dada.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def project(self, mean: np.ndarray, covariance: np.ndarray) -> tuple:
    """
    Project state distribution to measurement space.

    Args:
        mean (ndarray): The state's mean vector (8 dimensional array).
        covariance (ndarray): The state's covariance matrix (8x8 dimensional).

    Returns:
        (tuple[ndarray, ndarray]): Returns the projected mean and covariance matrix of the given state estimate.
    """
    std = [
        self._std_weight_position * mean[3],
        self._std_weight_position * mean[3],
        1e-1,
        self._std_weight_position * mean[3],
    ]
    innovation_cov = np.diag(np.square(std))

    mean = np.dot(self._update_mat, mean)
    covariance = np.linalg.multi_dot((self._update_mat, covariance, self._update_mat.T))
    return mean, covariance + innovation_cov

update(mean, covariance, measurement)

Ejecuta el paso de corrección del filtro Kalman.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio del estado predicho (8 dimensiones).

necesario
covariance ndarray

La matriz de covarianza del estado (8x8 dimensiones).

necesario
measurement ndarray

El vector de medida de 4 dimensiones (x, y, a, h), donde (x, y) es la posición central a la relación de aspecto y h la altura del cuadro delimitador.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve la distribución de estados corregida por la medición.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def update(self, mean: np.ndarray, covariance: np.ndarray, measurement: np.ndarray) -> tuple:
    """
    Run Kalman filter correction step.

    Args:
        mean (ndarray): The predicted state's mean vector (8 dimensional).
        covariance (ndarray): The state's covariance matrix (8x8 dimensional).
        measurement (ndarray): The 4 dimensional measurement vector (x, y, a, h), where (x, y) is the center
            position, a the aspect ratio, and h the height of the bounding box.

    Returns:
        (tuple[ndarray, ndarray]): Returns the measurement-corrected state distribution.
    """
    projected_mean, projected_cov = self.project(mean, covariance)

    chol_factor, lower = scipy.linalg.cho_factor(projected_cov, lower=True, check_finite=False)
    kalman_gain = scipy.linalg.cho_solve(
        (chol_factor, lower), np.dot(covariance, self._update_mat.T).T, check_finite=False
    ).T
    innovation = measurement - projected_mean

    new_mean = mean + np.dot(innovation, kalman_gain.T)
    new_covariance = covariance - np.linalg.multi_dot((kalman_gain, projected_cov, kalman_gain.T))
    return new_mean, new_covariance



ultralytics.trackers.utils.kalman_filter.KalmanFilterXYWH

Bases: KalmanFilterXYAH

Para BoT-SORT. Un sencillo filtro de Kalman para el seguimiento de cajas delimitadoras en el espacio de la imagen.

El espacio de estado de 8 dimensiones (x, y, w, h, vx, vy, vw, vh) contiene la posición central de la caja delimitadora (x, y), la anchura w, altura h, y sus respectivas velocidades.

El movimiento del objeto sigue un modelo de velocidad constante. La ubicación del cuadro delimitador (x, y, w, h) se toma como observación directa del espacio de estado (modelo de observación lineal).

Código fuente en ultralytics/trackers/utils/kalman_filter.py
class KalmanFilterXYWH(KalmanFilterXYAH):
    """
    For BoT-SORT. A simple Kalman filter for tracking bounding boxes in image space.

    The 8-dimensional state space (x, y, w, h, vx, vy, vw, vh) contains the bounding box center position (x, y), width
    w, height h, and their respective velocities.

    Object motion follows a constant velocity model. The bounding box location (x, y, w, h) is taken as direct
    observation of the state space (linear observation model).
    """

    def initiate(self, measurement: np.ndarray) -> tuple:
        """
        Create track from unassociated measurement.

        Args:
            measurement (ndarray): Bounding box coordinates (x, y, w, h) with center position (x, y), width, and height.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector (8 dimensional) and covariance matrix (8x8 dimensional)
                of the new track. Unobserved velocities are initialized to 0 mean.
        """
        mean_pos = measurement
        mean_vel = np.zeros_like(mean_pos)
        mean = np.r_[mean_pos, mean_vel]

        std = [
            2 * self._std_weight_position * measurement[2],
            2 * self._std_weight_position * measurement[3],
            2 * self._std_weight_position * measurement[2],
            2 * self._std_weight_position * measurement[3],
            10 * self._std_weight_velocity * measurement[2],
            10 * self._std_weight_velocity * measurement[3],
            10 * self._std_weight_velocity * measurement[2],
            10 * self._std_weight_velocity * measurement[3],
        ]
        covariance = np.diag(np.square(std))
        return mean, covariance

    def predict(self, mean, covariance) -> tuple:
        """
        Run Kalman filter prediction step.

        Args:
            mean (ndarray): The 8 dimensional mean vector of the object state at the previous time step.
            covariance (ndarray): The 8x8 dimensional covariance matrix of the object state at the previous time step.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
                velocities are initialized to 0 mean.
        """
        std_pos = [
            self._std_weight_position * mean[2],
            self._std_weight_position * mean[3],
            self._std_weight_position * mean[2],
            self._std_weight_position * mean[3],
        ]
        std_vel = [
            self._std_weight_velocity * mean[2],
            self._std_weight_velocity * mean[3],
            self._std_weight_velocity * mean[2],
            self._std_weight_velocity * mean[3],
        ]
        motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))

        mean = np.dot(mean, self._motion_mat.T)
        covariance = np.linalg.multi_dot((self._motion_mat, covariance, self._motion_mat.T)) + motion_cov

        return mean, covariance

    def project(self, mean, covariance) -> tuple:
        """
        Project state distribution to measurement space.

        Args:
            mean (ndarray): The state's mean vector (8 dimensional array).
            covariance (ndarray): The state's covariance matrix (8x8 dimensional).

        Returns:
            (tuple[ndarray, ndarray]): Returns the projected mean and covariance matrix of the given state estimate.
        """
        std = [
            self._std_weight_position * mean[2],
            self._std_weight_position * mean[3],
            self._std_weight_position * mean[2],
            self._std_weight_position * mean[3],
        ]
        innovation_cov = np.diag(np.square(std))

        mean = np.dot(self._update_mat, mean)
        covariance = np.linalg.multi_dot((self._update_mat, covariance, self._update_mat.T))
        return mean, covariance + innovation_cov

    def multi_predict(self, mean, covariance) -> tuple:
        """
        Run Kalman filter prediction step (Vectorized version).

        Args:
            mean (ndarray): The Nx8 dimensional mean matrix of the object states at the previous time step.
            covariance (ndarray): The Nx8x8 covariance matrix of the object states at the previous time step.

        Returns:
            (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
                velocities are initialized to 0 mean.
        """
        std_pos = [
            self._std_weight_position * mean[:, 2],
            self._std_weight_position * mean[:, 3],
            self._std_weight_position * mean[:, 2],
            self._std_weight_position * mean[:, 3],
        ]
        std_vel = [
            self._std_weight_velocity * mean[:, 2],
            self._std_weight_velocity * mean[:, 3],
            self._std_weight_velocity * mean[:, 2],
            self._std_weight_velocity * mean[:, 3],
        ]
        sqr = np.square(np.r_[std_pos, std_vel]).T

        motion_cov = [np.diag(sqr[i]) for i in range(len(mean))]
        motion_cov = np.asarray(motion_cov)

        mean = np.dot(mean, self._motion_mat.T)
        left = np.dot(self._motion_mat, covariance).transpose((1, 0, 2))
        covariance = np.dot(left, self._motion_mat.T) + motion_cov

        return mean, covariance

    def update(self, mean, covariance, measurement) -> tuple:
        """
        Run Kalman filter correction step.

        Args:
            mean (ndarray): The predicted state's mean vector (8 dimensional).
            covariance (ndarray): The state's covariance matrix (8x8 dimensional).
            measurement (ndarray): The 4 dimensional measurement vector (x, y, w, h), where (x, y) is the center
                position, w the width, and h the height of the bounding box.

        Returns:
            (tuple[ndarray, ndarray]): Returns the measurement-corrected state distribution.
        """
        return super().update(mean, covariance, measurement)

initiate(measurement)

Crea una pista a partir de una medición no asociada.

Par√°metros:

Nombre Tipo Descripción Por defecto
measurement ndarray

Coordenadas de la caja delimitadora (x, y, w, h) con posición central (x, y), anchura y altura.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media (8 dimensiones) y la matriz de covarianza (8x8 dimensiones) de la nueva pista. Las velocidades no observadas se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def initiate(self, measurement: np.ndarray) -> tuple:
    """
    Create track from unassociated measurement.

    Args:
        measurement (ndarray): Bounding box coordinates (x, y, w, h) with center position (x, y), width, and height.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector (8 dimensional) and covariance matrix (8x8 dimensional)
            of the new track. Unobserved velocities are initialized to 0 mean.
    """
    mean_pos = measurement
    mean_vel = np.zeros_like(mean_pos)
    mean = np.r_[mean_pos, mean_vel]

    std = [
        2 * self._std_weight_position * measurement[2],
        2 * self._std_weight_position * measurement[3],
        2 * self._std_weight_position * measurement[2],
        2 * self._std_weight_position * measurement[3],
        10 * self._std_weight_velocity * measurement[2],
        10 * self._std_weight_velocity * measurement[3],
        10 * self._std_weight_velocity * measurement[2],
        10 * self._std_weight_velocity * measurement[3],
    ]
    covariance = np.diag(np.square(std))
    return mean, covariance

multi_predict(mean, covariance)

Ejecuta el paso de predicción del filtro Kalman (versión vectorizada).

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

La matriz media de Nx8 dimensiones de los estados del objeto en el paso de tiempo anterior.

necesario
covariance ndarray

La matriz de covarianza Nx8x8 de los estados del objeto en el paso de tiempo anterior.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media y la matriz de covarianza del estado predicho. Las velocidades no observadas de se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def multi_predict(self, mean, covariance) -> tuple:
    """
    Run Kalman filter prediction step (Vectorized version).

    Args:
        mean (ndarray): The Nx8 dimensional mean matrix of the object states at the previous time step.
        covariance (ndarray): The Nx8x8 covariance matrix of the object states at the previous time step.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
            velocities are initialized to 0 mean.
    """
    std_pos = [
        self._std_weight_position * mean[:, 2],
        self._std_weight_position * mean[:, 3],
        self._std_weight_position * mean[:, 2],
        self._std_weight_position * mean[:, 3],
    ]
    std_vel = [
        self._std_weight_velocity * mean[:, 2],
        self._std_weight_velocity * mean[:, 3],
        self._std_weight_velocity * mean[:, 2],
        self._std_weight_velocity * mean[:, 3],
    ]
    sqr = np.square(np.r_[std_pos, std_vel]).T

    motion_cov = [np.diag(sqr[i]) for i in range(len(mean))]
    motion_cov = np.asarray(motion_cov)

    mean = np.dot(mean, self._motion_mat.T)
    left = np.dot(self._motion_mat, covariance).transpose((1, 0, 2))
    covariance = np.dot(left, self._motion_mat.T) + motion_cov

    return mean, covariance

predict(mean, covariance)

Ejecuta el paso de predicción del filtro Kalman.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio de 8 dimensiones del estado del objeto en el paso de tiempo anterior.

necesario
covariance ndarray

La matriz de covarianza de 8x8 dimensiones del estado del objeto en el paso de tiempo anterior.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve el vector media y la matriz de covarianza del estado predicho. Las velocidades no observadas de se inicializan con media 0.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def predict(self, mean, covariance) -> tuple:
    """
    Run Kalman filter prediction step.

    Args:
        mean (ndarray): The 8 dimensional mean vector of the object state at the previous time step.
        covariance (ndarray): The 8x8 dimensional covariance matrix of the object state at the previous time step.

    Returns:
        (tuple[ndarray, ndarray]): Returns the mean vector and covariance matrix of the predicted state. Unobserved
            velocities are initialized to 0 mean.
    """
    std_pos = [
        self._std_weight_position * mean[2],
        self._std_weight_position * mean[3],
        self._std_weight_position * mean[2],
        self._std_weight_position * mean[3],
    ]
    std_vel = [
        self._std_weight_velocity * mean[2],
        self._std_weight_velocity * mean[3],
        self._std_weight_velocity * mean[2],
        self._std_weight_velocity * mean[3],
    ]
    motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))

    mean = np.dot(mean, self._motion_mat.T)
    covariance = np.linalg.multi_dot((self._motion_mat, covariance, self._motion_mat.T)) + motion_cov

    return mean, covariance

project(mean, covariance)

Proyecta la distribución de estados al espacio de medida.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio del estado (matriz de 8 dimensiones).

necesario
covariance ndarray

La matriz de covarianza del estado (8x8 dimensiones).

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve la media proyectada y la matriz de covarianza de la estimación de estado dada.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def project(self, mean, covariance) -> tuple:
    """
    Project state distribution to measurement space.

    Args:
        mean (ndarray): The state's mean vector (8 dimensional array).
        covariance (ndarray): The state's covariance matrix (8x8 dimensional).

    Returns:
        (tuple[ndarray, ndarray]): Returns the projected mean and covariance matrix of the given state estimate.
    """
    std = [
        self._std_weight_position * mean[2],
        self._std_weight_position * mean[3],
        self._std_weight_position * mean[2],
        self._std_weight_position * mean[3],
    ]
    innovation_cov = np.diag(np.square(std))

    mean = np.dot(self._update_mat, mean)
    covariance = np.linalg.multi_dot((self._update_mat, covariance, self._update_mat.T))
    return mean, covariance + innovation_cov

update(mean, covariance, measurement)

Ejecuta el paso de corrección del filtro Kalman.

Par√°metros:

Nombre Tipo Descripción Por defecto
mean ndarray

El vector medio del estado predicho (8 dimensiones).

necesario
covariance ndarray

La matriz de covarianza del estado (8x8 dimensiones).

necesario
measurement ndarray

El vector de medida de 4 dimensiones (x, y, w, h), donde (x, y) es la posición central w la anchura y h la altura de la caja delimitadora.

necesario

Devuelve:

Tipo Descripción
tuple[ndarray, ndarray]

Devuelve la distribución de estados corregida por la medición.

Código fuente en ultralytics/trackers/utils/kalman_filter.py
def update(self, mean, covariance, measurement) -> tuple:
    """
    Run Kalman filter correction step.

    Args:
        mean (ndarray): The predicted state's mean vector (8 dimensional).
        covariance (ndarray): The state's covariance matrix (8x8 dimensional).
        measurement (ndarray): The 4 dimensional measurement vector (x, y, w, h), where (x, y) is the center
            position, w the width, and h the height of the bounding box.

    Returns:
        (tuple[ndarray, ndarray]): Returns the measurement-corrected state distribution.
    """
    return super().update(mean, covariance, measurement)





Creado 2023-11-12, Actualizado 2024-05-18
Autores: glenn-jocher (4), Burhan-Q (1)