Sensor-Image Synchronization for Sports Motion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately detecting important moments during continuous sports motions from images, as the duration of these moments is short and involves complex motions, making it difficult to correctly capture them.
Innovation Solution
A system comprising a sensor device, smartphone, and camera that synchronizes sensor data with image timestamps to identify and extract important moments, allowing for precise selection and recording of still images or moving images based on detected motion patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If expression or action detection technology is applied to identify important moments, then the system can automatically detect triggers from images, but the detection accuracy deteriorates for short-duration continuous motions in sports
Solution Approach 1:
The patent segments the detection task into two parts: using sensor data (acceleration, angular velocity) to detect motion events and using image data to capture and verify the moments. The sensor data is processed in segments (before, during, after the detected motion event) to identify important moments with high precision while maintaining automation.
Solution Approach 2:
The patent introduces sensor data as an intermediary between the camera imaging system and the important moment detection. The sensor device detects motion events and provides timing information that guides the image capture and selection process, enabling accurate detection of short-duration sports motions while maintaining automatic operation.
2Reliability
If continuous imaging is performed to capture sports motions, then all motion details are recorded, but the difficulty of identifying important moments increases due to continuous complex motions
Solution Approach 1:
The patent performs preliminary detection of motion events using sensor data before finalizing the important moment identification. The sensor device continuously monitors acceleration and angular velocity, and when a motion event is detected, the system prepares to capture and analyze corresponding images, making the identification process easier and more reliable.
Solution Approach 2:
The patent implements a feedback loop where sensor data is continuously fed back to guide image capture and selection. The detected motion events from sensor data provide feedback that adjusts the imaging parameters and selection criteria, enabling reliable capture of important moments while reducing the difficulty of identification through continuous optimization.
3Measurement precision
If sensor data is synchronized with image timestamps to identify important moments, then the precision of moment extraction is improved, but the system complexity increases
Solution Approach 1:
The patent merges the sensor data processing and image processing systems into a unified synchronization framework. Both data streams are timestamped and processed together, allowing precise moment extraction while reducing overall system complexity through integration rather than separate complex subsystems.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A non-transitory computer readable storage device includes a storage medium having a downloaded application stored therein. The downloaded application has instructions that when executed by processing circuitry configure the processing circuitry to receive sensor data from a sensor. The sensor is attached to a person or attached to an item used by the person. The processing circuitry also analyzes a motion pattern in the sensor data to identify a predetermined event in the motion pattern, the predetermined event captured in an image or series of images by an image capture device.