create_test_data.py

"""Generate a synthetic test video and GPS JSON for testing the pipeline."""

import json
from datetime import datetime, timedelta

import cv2
import numpy as np


def create_test_video(output_path: str, fps: int = 30, duration_sec: int = 10):
    """Create a simple test video simulating a road with a pothole."""
    width, height = 640, 480
    fourcc = cv2.VideoWriter_fourcc(*"mp4v")
    writer = cv2.VideoWriter(output_path, fourcc, fps, (width, height))

    total_frames = fps * duration_sec

    for i in range(total_frames):
        # Grey road background
        frame = np.full((height, width, 3), (100, 100, 100), dtype=np.uint8)

        # Draw road lines (white dashes)
        for y in range(0, height, 40):
            offset = (i * 3) % 40  # animate movement
            cv2.line(frame, (width // 2, y - offset), (width // 2, y - offset + 20), (255, 255, 255), 2)

        # Draw road edges
        cv2.line(frame, (100, 0), (100, height), (200, 200, 200), 3)
        cv2.line(frame, (540, 0), (540, height), (200, 200, 200), 3)

        # Between frames 90-210 (seconds 3-7): draw a dark pothole-like shape
        if 90 <= i <= 210:
            center_x = 300
            center_y = 300
            axes = (40, 25)
            cv2.ellipse(frame, (center_x, center_y), axes, 0, 0, 360, (30, 30, 30), -1)
            # Add rough edges to make it look more like a pothole
            cv2.ellipse(frame, (center_x, center_y), (45, 30), 0, 0, 360, (50, 50, 50), 2)
            # Add text label for clarity
            cv2.putText(frame, "POTHOLE", (center_x - 40, center_y - 40),
                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1)

        # Between frames 240-300 (seconds 8-10): draw debris on road
        if 240 <= i <= 300:
            # Random-ish debris shapes
            pts = np.array([[350, 200], [380, 190], [390, 210], [370, 220], [345, 215]], np.int32)
            cv2.fillPoly(frame, [pts], (0, 80, 150))
            cv2.putText(frame, "DEBRIS", (340, 180),
                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1)

        # Add frame number
        cv2.putText(frame, f"Frame {i}", (10, 30),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 1)

        writer.write(frame)

    writer.release()
    print(f"Test video created: {output_path} ({total_frames} frames, {duration_sec}s)")


def create_test_gps_json(output_path: str, duration_sec: int = 10):
    """Create a GPS JSON file with ~3 entries per second."""
    start_time = datetime(2026, 3, 21, 10, 0, 0)
    start_lat = 40.7128
    start_lon = -74.0060

    entries = []
    intervals_per_sec = 3  # multiple entries per second

    for i in range(duration_sec * intervals_per_sec):
        t = start_time + timedelta(milliseconds=i * (1000 // intervals_per_sec))
        # Simulate slow movement north
        lat = start_lat + (i * 0.00001)
        lon = start_lon + (i * 0.000005)
        entries.append({
            "timestamp": t.isoformat(),
            "gps_lat": round(lat, 7),
            "gps_lon": round(lon, 7),
        })

    with open(output_path, "w") as f:
        json.dump(entries, f, indent=2)

    print(f"GPS JSON created: {output_path} ({len(entries)} entries)")


if __name__ == "__main__":
    create_test_video("test_video.mp4")
    create_test_gps_json("test_gps.json")