Video Tracking Correction via Selective Frame Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving body tracking systems, such as those in video tracking devices, often incorrectly track a moving body like a soccer ball due to similarities in color or shape with other objects, leading to errors and increased workload for confirmation and correction.
Innovation Solution
A correcting and verifying method and device that displays specific frame images for user confirmation and correction, using a processor to detect ball candidates, track the ball, and correct its position based on user instructions, reducing the workload by allowing selective verification and correction of tracking results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If automatic tracking is performed using color and shape features, then tracking speed is improved, but tracking accuracy deteriorates due to misidentification with similar objects
Solution Approach 1:
The system displays frame images where tracking errors are likely to occur and receives user feedback on correct positions. This feedback is used to correct tracking results and improve future tracking accuracy, resolving the contradiction between fast automatic tracking and accurate identification.
Solution Approach 2:
The system introduces an intermediary verification step where user input serves as a mediator between automatic tracking and final results. This allows the system to maintain fast automatic tracking while periodically correcting errors through user feedback on specific frames.
2Measurement precision
If all frame images are manually verified, then tracking accuracy is improved, but work load increases significantly
Solution Approach 1:
The system extracts and displays only specific frame images where tracking errors are likely to occur, rather than requiring verification of all frames. This reduces the workload from manual verification of every frame to targeted verification of critical frames only.
Solution Approach 2:
Instead of requiring complete manual verification of all frames, the system performs partial verification on selected frames where errors are most likely. This partial action is sufficient to maintain high tracking accuracy without the excessive workload of full verification.
3Adaptability or versatility
If tracking is performed on similar objects like ball and foot, then tracking coverage is improved, but error rate increases due to visual similarity
Solution Approach 1:
The system uses user feedback on frames where similar objects appear to learn from mistakes and improve discrimination between visually similar objects like balls and feet, maintaining high tracking coverage while reducing error rates.
Solution Approach 2:
The system performs preliminary tracking on all objects including similar ones, then uses feedback from critical frames to correct errors before final output. This allows comprehensive tracking coverage while maintaining reliability through corrective action.
Data Source
AI summary
A correcting and verifying method and a correcting and verifying device cause a processor to display a specific frame image as a frame image to be confirmed based on a tracking result of a moving body in each of a plurality of frame images which configure a video, and to correct a position of the moving body in the frame image to be confirmed in a case where a correction instruction of a user is received.