Camera Calibration

Camera calibration is essential for accurate biomechanical analysis. It establishes the relationship between camera pixels and real-world measurements.

Requirements

You'll need a checkerboard calibration pattern with known dimensions. Common configurations:

• 4×5 checkerboard (4 rows, 5 columns)
• 35mm square size
• Printed on stiff material (foam board recommended)

Calibration Process

1. Create a Session

First, create a calibration session:

Swift:

let session = try await service.createSession()

TypeScript:

const session = await service.createSession();

2. Set Up Checkerboard

Define your checkerboard details:

Swift:

let details = CheckerboardDetails(
    rows: 4,
    columns: 5,
    squareSize: 35,  // millimeters
    placement: .perpendicular
)

TypeScript:

const details = {
  rows: 4,
  columns: 5,
  squareSize: 35,  // millimeters
  placement: "perpendicular"
};

3. Calibrate Camera

Swift:

try await service.calibrateCamera(session, checkerboardDetails: details) { status in
    switch status {
    case .detectingCheckerboard:
        print("Looking for checkerboard...")
    case .capturedImage(let count, let required):
        print("Captured \(count)/\(required) images")
    case .completed:
        print("Calibration complete!")
    }
}

TypeScript:

await service.calibrateCamera(session.id, details, (status) => {
  if (status.type === "detecting") {
    console.log("Looking for checkerboard...");
  } else if (status.type === "captured") {
    console.log(`Captured ${status.count}/${status.required} images`);
  } else if (status.type === "completed") {
    console.log("Calibration complete!");
  }
});

Neutral Pose Calibration

After camera calibration, capture the subject's neutral standing pose for scaling:

Swift:

let subjects = try await service.subjectList()
if let subject = subjects.first {
    try await service.calibrateNeutralPose(for: subject, in: session) { status in
        switch status {
        case .detectingPose:
            print("Detecting neutral pose...")
        case .completed:
            print("Neutral pose captured!")
        }
    }
}

TypeScript:

const subjects = await service.subjectList();
if (subjects.length > 0) {
  await service.calibrateNeutralPose(subjects[0].id, session.id, (status) => {
    if (status === "detecting") {
      console.log("Detecting neutral pose...");
    } else if (status === "completed") {
      console.log("Neutral pose captured!");
    }
  });
}

Tips for Successful Calibration

1. Lighting - Ensure even, bright lighting without glare on the checkerboard
2. Movement - Move the checkerboard to various positions and angles
3. Coverage - Cover different areas of the camera's field of view
4. Stillness - Hold the checkerboard still when capturing each image
5. Distance - Vary the distance from the camera during calibration

Troubleshooting

Checkerboard not detected:

• Ensure the pattern is clearly visible
• Check lighting conditions
• Verify the rows/columns configuration matches your physical checkerboard

Poor calibration quality:

• Capture more images from different angles
• Ensure the checkerboard is flat and not warped
• Try different distances from the camera