Loading / 加载中...
Loading / 加载中...

Above-Water Stitching + Underwater Stitching + AI Swimming Stroke Recognition · Building a Unified Platform for Event Panoramic Monitoring and Coaching Analysis
Designed for competitive swimming and synchronized swimming events, the system stitches multiple scattered above-water and underwater camera feeds into two continuous panoramic video streams. Combined with AI swimming stroke recognition and athlete tracking, it provides an integrated visual solution for live streaming, referee judgment, coaching review, and technical instruction.
Industry Pain Points · Industry Pain Points
With a 50-meter pool, a single camera can only cover a portion of the lanes. Above-water and underwater footage are scattered across multiple windows, making it impossible for judges and directors to maintain full situational awareness.
Traditional surveillance focuses on surface-level views, making it difficult to capture underwater turns, wall pushes, starts, and swimming technique details clearly, affecting referee judgment and technical analysis.
Post-event technical analysis requires searching multiple recorded video segments. Coaches struggle to quickly locate an athlete's complete swimming lap, reducing coaching efficiency.
In synchronized swimming events, athletes' formations, synchronized movements, and underwater postures require multi-angle simultaneous evaluation. Current multi-screen solutions fail to support precise scoring.
Aquatic Event Vision
The core challenge of swimming events lies in the large pool span, fragmented above-water and underwater perspectives, and complex synchronized swimming evaluation dimensions. Using video stitching as the foundation, the system separately fuses above-water and underwater cameras into two independent panoramic video streams. Combined with AI swimming stroke recognition and athlete tracking capabilities, it enables event monitoring, live direction, referee judgment, and coaching review all within a unified platform.
Real-time stitching of multiple cameras on both sides and ends of the pool into a continuous panoramic video, completely capturing starts, turns, and sprint finishes across all lanes.
Integrating pool bottom and side-wall underwater cameras into one panoramic underwater video stream, clearly capturing wall pushes, starts, turns, and underwater swimming technique details.
Real-time skeletal keypoint detection on the panoramic video stream automatically identifies freestyle, butterfly, backstroke, and breaststroke techniques, output motion scoring and technical deviation annotations.
Panoramic recordings support rapid retrieval by athlete, lane, or time segment. Coaches can compare multiple athletes' technical movements on the same screen, improving coaching efficiency.


Above-water and underwater stitching operate independently, can be deployed separately, or output dual panoramic video streams simultaneously, accommodating different event scales and budgets.
Plan One
Ideal for prioritizing event direction, referee monitoring, and large-screen display needs. The system fuses multiple cameras on both sides and ends of the pool in real-time into a continuous panoramic video, completely covering all lanes' starts, turns, sprint finishes, and touch-wall moments. It supports overlaying athlete numbers, lane markings, and timing data.

Plan Two
Ideal for events requiring precise technical analysis, referee judgment assistance, and synchronized swimming scoring. The system integrates multiple underwater cameras from the pool bottom and side walls into one panoramic underwater video stream, clearly presenting wall pushes, starts, turns, underwater swimming techniques, and synchronized swimming underwater lifting movements, with AI recognition outputting technical evaluation data.

AI-Powered Analysis
By overlaying AI analysis on the panoramic video stream, the system achieves the leap from 'watching footage' to 'understanding movements'. It can output real-time swimming stroke types, motion keypoints, formation structures, and pacing curves, providing quantified technical feedback to coaches, judges, and athletes while supporting multi-dimensional synchronized swimming scoring assistance.
Real-time detection of freestyle, butterfly, backstroke, and breaststroke, outputting each athlete's stroke type and movement rhythm.
Extracting critical technical metrics such as arm entry angle, stroke frequency, leg kick amplitude, and turn timing, generating quantified assessment reports.
Identifying multi-athlete formation structures, synchronized movement deviations, and underwater lifting postures to assist referee scoring and technical coaching.
Continuously tracking each athlete's position and velocity across the panoramic view, automatically generating segment-by-segment pacing curves.
| Feature | Supported |
|---|---|
| Real-time stitching of multiple above-water cameras into one panoramic video | ✓ |
| Real-time stitching of multiple underwater cameras into one panoramic video | ✓ |
| Synchronized display and recording of above-water and underwater panoramic views | ✓ |
| Simultaneous viewing of original multi-camera feeds and stitched panoramic view | ✓ |
| AI automatic recognition of four competitive swimming techniques and motion keypoints | ✓ |
| Formation structure and synchronized movement deviation analysis for synchronized swimming | ✓ |
| Athlete lane tracking and segment-level pacing data output | ✓ |
| Post-event panoramic replay search by athlete / lane / time segment | ✓ |
| Teaching replay, motion comparison, and technical annotation overlay support | ✓ |
| Live event direction switching, large-screen display, and mobile viewing support | ✓ |
One above-water panoramic video covers all lanes, eliminating blind spots for directors and judges
One underwater panoramic video completely records wall pushes, starts, and underwater swimming techniques, supporting precise judgment
AI swimming stroke recognition multiplies manual technical analysis efficiency by several times, reducing coaching review costs
Synchronized swimming events benefit from simultaneous above-water formation and underwater posture display, supporting multi-dimensional scoring
Panoramic recordings and AI annotation data accumulate as training archives, supporting long-term technical tracking
Reuses existing pool camera infrastructure, progressive upgrades protect existing investment
Supporting pool location assessment, above/below-water stitching debugging, AI swimming stroke recognition integration, and event coaching platform integration
Related Products