← Back to list

video-analysis-workflow
by feniix
Video-based kinematic analysis for athletic performance.
⭐ 0🍴 0📅 Jan 20, 2026
SKILL.md
name: video-analysis-workflow description: Guides video analysis for CMJ and drop jump. Use when processing athlete videos, debugging pose detection, troubleshooting analysis failures, or running kinemotion CLI commands.
Video Analysis Workflow
Analysis Commands
# CMJ analysis
uv run kinemotion cmj-analyze <video> [--output debug.mp4]
# Drop Jump analysis
uv run kinemotion dropjump-analyze <video> [--output debug.mp4]
# Batch processing
uv run kinemotion cmj-analyze videos/*.mp4 --batch --workers 4
Quality Presets
| Preset | Use Case | Trade-off |
|---|---|---|
fast | Quick preview, large batches | Lower accuracy |
balanced | Default, most use cases | Good accuracy/speed |
accurate | Validation, research | Best accuracy, slower |
uv run kinemotion cmj-analyze video.mp4 --quality accurate
Debug Output
Always use --output debug.mp4 when:
- Metrics seem incorrect
- Troubleshooting pose detection
- Validating new videos
- Training coaches on video quality
The debug video shows:
- Skeleton overlay with joint angles
- Phase markers (takeoff, landing, peak)
- Real-time metrics display
Camera Angle Recommendations
| Angle | Recommendation | Reason |
|---|---|---|
| 45° oblique | Recommended | Both legs clearly visible, accurate tracking |
| 90° lateral | Not recommended | MediaPipe confuses left/right feet (occlusion) |
| Front/back | Not recommended | Depth ambiguity for sagittal plane motion |
Troubleshooting
No Takeoff Detected
- Verify video contains complete jump (before, during, after)
- Check athlete is fully visible throughout
- Try
--quality accuratefor stricter detection - Review debug video for landmark quality
Invalid Metrics
- Video may be too short (need full jump cycle)
- Athlete may be partially occluded
- Poor lighting affecting pose detection
- Camera shake causing landmark jitter
Jittery Landmarks
- Check lighting conditions (avoid backlighting)
- Ensure stable camera (tripod recommended)
- Verify athlete clothing contrast with background
- Try
--quality accuratefor better filtering
Rotation Issues (Mobile Videos)
- Mobile videos often have rotation metadata
- kinemotion handles this automatically via
video_io.py - If issues persist, pre-process with:
ffmpeg -i input.mp4 -vf "transpose=1" output.mp4
Video Requirements
| Requirement | Specification |
|---|---|
| Frame rate | 30+ fps (60+ preferred) |
| Resolution | 720p minimum |
| Duration | Full jump cycle visible |
| Lighting | Even, front-lit preferred |
| Background | Contrasting with athlete |
| Camera | Stable, tripod recommended |
Output Metrics
CMJ Metrics
jump_height_cm: Calculated from flight timeflight_time_ms: Time in aircountermovement_depth_cm: Lowest point before takeofftakeoff_velocity_m_s: Velocity at ground leavetriple_extension: Hip, knee, ankle angles at takeoff
Drop Jump Metrics
ground_contact_time_ms: Time on ground after dropflight_time_ms: Time in air after contactreactive_strength_index: RSI = flight_time / contact_timedrop_height_cm: Initial drop height (if detectable)
Python API Alternative
from kinemotion import process_cmj_video, process_dropjump_video
# CMJ
metrics = process_cmj_video("video.mp4", quality="balanced")
# Drop Jump
metrics = process_dropjump_video("video.mp4", quality="balanced")
Advanced Troubleshooting
If you encounter issues not covered here:
- Use
ref_search_documentationto check Kinemotion docs or MediaPipe/OpenCV docs. - Use
web_search_exato search for similar issues online.
Score
Total Score
60/100
Based on repository quality metrics
✓SKILL.md
SKILL.mdファイルが含まれている
+20
✓LICENSE
ライセンスが設定されている
+10
○説明文
100文字以上の説明がある
0/10
○人気
GitHub Stars 100以上
0/15
○最近の活動
3ヶ月以内に更新がある
0/10
○フォーク
10回以上フォークされている
0/5
✓Issue管理
オープンIssueが50未満
+5
✓言語
プログラミング言語が設定されている
+5
○タグ
1つ以上のタグが設定されている
0/5
Reviews
💬
Reviews coming soon