| """Trajectory rendering helpers.""" |
|
|
| from __future__ import annotations |
|
|
| from collections.abc import Sequence |
|
|
| import cv2 |
| import numpy as np |
|
|
|
|
| def draw_trajectory( |
| frame: np.ndarray, |
| points: Sequence[tuple[float, float]], |
| color: tuple[int, int, int], |
| thickness: int = 2, |
| ) -> np.ndarray: |
| """Draw a motion trail on a frame.""" |
|
|
| if len(points) < 2: |
| return frame |
|
|
| int_points = [(int(x), int(y)) for x, y in points] |
| for start, end in zip(int_points[:-1], int_points[1:], strict=False): |
| cv2.line(frame, start, end, color, thickness, lineType=cv2.LINE_AA) |
| return frame |
|
|
|
|
| def relative_speed(points: Sequence[tuple[int, float, float]]) -> float: |
| """Estimate average pixel displacement per frame for a trajectory.""" |
|
|
| if len(points) < 2: |
| return 0.0 |
|
|
| speeds: list[float] = [] |
| for (frame_a, x_a, y_a), (frame_b, x_b, y_b) in zip(points[:-1], points[1:], strict=False): |
| frame_delta = max(frame_b - frame_a, 1) |
| distance = float(np.hypot(x_b - x_a, y_b - y_a)) |
| speeds.append(distance / frame_delta) |
| return float(np.mean(speeds)) if speeds else 0.0 |
|
|
|
|
