Motion tracking is the process of identifying and following specific anatomical landmarks or points on the body as they move through space during a video recording. This technique forms the foundation of quantitative video analysis, enabling measurement of positions, angles, velocities, and other biomechanical variables.
Types of Motion Tracking
Manual Tracking
- Analyst manually marks points frame-by-frame
- Most accurate for complex or partially obscured movements
- Time-consuming but highly reliable
- Allows for correction and quality control
- Ideal when automated methods fail
Semi-Automated Tracking
- Software suggests point locations, analyst confirms or adjusts
- Balances accuracy with efficiency
- Reduces analyst fatigue
- Common in professional motion analysis systems
- Good for most clinical and coaching applications
Fully Automated Tracking
- Software automatically identifies and tracks points
- Fastest method, suitable for high-volume analysis
- Requires clear visibility and good contrast
- May need post-processing corrections
- Increasingly powered by machine learning and AI
Marker-Based Tracking
- Physical markers (reflective, colored, or contrast markers) placed on body
- Easier for software to track automatically
- Common in research laboratories
- Provides highest accuracy
- Requires preparation time and marker placement
Markerless Tracking
- Tracks anatomical landmarks directly without physical markers
- More practical for field use and coaching
- Modern AI methods improving accuracy significantly
- No preparation needed
- Ideal for retrospective analysis of existing video
Common Tracking Points
Lower body analysis typically includes:
- Ankle (lateral malleolus)
- Knee (lateral femoral condyle)
- Hip (greater trochanter)
- Pelvis (iliac crest or ASIS)
- Toe and heel
Upper body analysis may include:
- Wrist
- Elbow
- Shoulder (acromion)
- Head
- Spine (various vertebrae)
Data Generated from Motion Tracking
Position Data
- X-Y coordinates (2D analysis) or X-Y-Z coordinates (3D analysis)
- Trajectory paths showing movement patterns
- Displacement values for each point
Derived Measurements
- Joint Angles: Calculated from three tracked points forming an angle
- Velocities: Change in position over time
- Accelerations: Change in velocity over time
- Segment Lengths: Distances between tracked points
- Center of Mass: Calculated from multiple segment positions
Factors Affecting Tracking Quality
Video Quality
- Resolution: Higher resolution enables more precise point identification
- Frame rate: Higher rates provide more data points
- Contrast: Clear visibility of landmarks or markers
- Lighting: Even, adequate illumination without shadows
- Focus: Sharp, non-blurry images
Camera Setup
- Perpendicular view to plane of motion for 2D analysis
- Stable, non-moving camera position
- Appropriate distance from subject
- Correct camera height
- Multiple synchronized cameras for 3D analysis
Movement Characteristics
- Speed: Faster movements may blur even with high shutter speeds
- Plane of motion: Movements in single plane easier to track
- Occlusion: Body parts blocking view of landmarks
- Clothing: Tight-fitting clothes or markers needed for accuracy
Tracking Workflow
1. Video Preparation: - Import video into analysis software - Set appropriate playback settings - Define region of interest
2. Calibration: - Establish real-world scale (see Calibration term) - Set coordinate system - Define measurement units
3. Point Identification: - Identify anatomical landmarks - Create tracking points - Name points systematically
4. Tracking Process: - Track points through entire video or region of interest - Review automatically tracked data - Correct errors or missing frames
5. Data Processing: - Filter noise if necessary - Calculate derived variables (angles, velocities) - Export data for further analysis
6. Visualization: - Create stick figures or overlay graphics - Generate trajectory plots - Produce angle-time curves or other visualizations
Applications
Clinical Gait Analysis
- Track ankle, knee, and hip through gait cycle
- Calculate joint angles at key events
- Assess symmetry between limbs
- Monitor rehabilitation progress
Sports Performance
- Analyze throwing, kicking, or striking mechanics
- Track implement path (bat, racquet, club)
- Measure segment velocities for power assessment
- Compare technique before and after coaching interventions
Ergonomics
- Track joint positions during work tasks
- Assess repetitive strain risk
- Design workspace improvements
- Evaluate intervention effectiveness
Modern Advances
AI and Deep Learning
- Convolutional neural networks for automatic point detection
- Improved accuracy for markerless tracking
- Real-time tracking becoming practical
- Handling of partial occlusions
Mobile Applications
- Smartphone apps with built-in tracking
- Accessible to coaches and athletes
- Immediate feedback
- Lower cost than traditional systems
3D Markerless Systems
- Multiple camera angles
- Depth sensors
- Full body tracking without markers
- Movement capture for animation and detailed analysis
Motion tracking transforms qualitative video into quantitative data, enabling objective assessment, precise measurement, and scientific analysis of human movement. As technology improves, tracking becomes more accurate, faster, and accessible to a wider range of users.