Skip to contents

Foot-IMU spatiotemporal gait parameters

Source: R/imu-gait.R
footImuGait.Rd

End-to-end foot-mounted IMU gait pipeline: estimate orientation, remove gravity, rotate acceleration into the world frame, detect stance, and ZUPT-integrate to a drift-bounded foot trajectory, then derive per-stride stride length, foot clearance, stance/swing time and gait velocity.

Usage

footImuGait(
  accel,
  gyro,
  sampling_rate,
  g = 9.81,
  orientation = NULL,
  orientation_method = c("madgwick", "complementary"),
  beta = 0.02,
  stance = NULL,
  vertical_axis = 3,
  ...
)

Arguments

accel

Numeric matrix (n x 3) of accelerometer readings (m/s^2).

gyro

Numeric matrix (n x 3) of gyroscope readings (rad/s).

sampling_rate

Sampling rate in Hz.

g

Gravitational acceleration magnitude (default 9.81 m/s^2).

orientation

Optional orientation (a data frame with q_w/q_x/q_y/ q_z, or an n x 4 quaternion matrix). If NULL, estimateOrientation() is run internally.

orientation_method

Sensor-fusion method for the internal orientation estimate, "madgwick" (default) or "complementary". Ignored when orientation is supplied.

beta

Madgwick filter gain for the internal orientation estimate (default 0.02). A foot IMU sees large dynamic accelerations, so a small, gyro-dominant gain avoids swing-phase tilt error; raise it for lower-dynamic mounting or noisier gyroscopes. Ignored when orientation is supplied.

stance

Optional logical stance vector; if NULL, detectStanceZUPT() is used.

vertical_axis

World axis that points up (1, 2 or 3; default 3). Used for foot clearance and to de-drift vertical position to the floor.

...

Additional arguments forwarded to detectStanceZUPT() (e.g. method, accel_threshold).

Value

An imu_gait object: a list with strides (a data frame with one row per detected stride: stride, stride_length, foot_clearance, stance_time, swing_time, stride_time, gait_velocity), the world-frame position and velocity matrices, the stance mask, and sampling_rate.

References

Mariani B et al. (2010) J Biomech 43(15):2999-3006; Rebula JR, Ojeda LV, Adamczyk PG, Kuo AD (2013). "Measurement of foot placement and its variability with inertial sensors." Gait & Posture 38(4):974-980.

See also

detectStanceZUPT(), strapdownIntegrate(), estimateOrientation(), removeGravity().

Examples

set.seed(1)
g <- 9.81
stance_block <- matrix(rep(c(0, 0, -g), each = 60), 60, 3)
swing_block <- matrix(rnorm(40 * 3, 0, 3), 40, 3)
accel <- rbind(stance_block, swing_block, stance_block, swing_block, stance_block)
gyro <- rbind(matrix(0, 60, 3), matrix(rnorm(40 * 3, 0, 1.5), 40, 3),
              matrix(0, 60, 3), matrix(rnorm(40 * 3, 0, 1.5), 40, 3),
              matrix(0, 60, 3))
res <- footImuGait(accel, gyro, sampling_rate = 200)
names(res)
#> [1] "strides"       "position"      "velocity"      "stance"       
#> [5] "sampling_rate"