Computer Vision Asset Tracking Linking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing tracking systems face inaccuracies and inefficiencies in linking tracking devices with tracked assets due to user errors, incomplete information, and overhead in manual input processes, particularly when dealing with large numbers of assets.
Innovation Solution
Implementing an image or video-based computer vision technique to automatically identify and associate tracking devices with tracked assets, using visual identifiers, radio signals, and machine learning models to improve accuracy and reduce manual input requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual input processes are used to link tracking devices with tracked assets, then user control and flexibility are maintained, but user errors and incomplete information occur leading to inaccurate link information
Solution Approach 1:
The system performs automatic identification and association of tracking devices with tracked assets using computer vision techniques. The image processing system independently captures images, identifies assets and tracking devices, and establishes links without requiring manual user input, thereby eliminating user errors while maintaining operational simplicity
Solution Approach 2:
The patent replaces manual mechanical input processes with an automated optical system. Instead of users manually entering data to link tracking devices with assets, the system uses image capture and computer vision algorithms to automatically identify and associate the two, substituting human operation with an automated imaging and processing system
2Productivity
If manual input processes are used for linking tracking devices and assets, then system complexity remains low, but overhead increases particularly when dealing with large numbers of assets
Solution Approach 1:
The system uses image copies (photographs) of the physical environment to identify and associate tracking devices with tracked assets. By processing visual representations rather than requiring direct physical interaction with each asset, the system efficiently handles large numbers of assets simultaneously while maintaining manageable computational complexity through standard image processing techniques
3Measurement precision
If automatic image-based identification is implemented, then user input errors are reduced and accuracy improves, but system complexity and processing requirements increase
Solution Approach 1:
The patent introduces an image as an intermediary between the tracking device and the tracked asset. Instead of directly linking devices to assets through complex protocols, the system captures an image that contains both elements, uses the image as a mediator to extract identification information, and establishes the association. This intermediary approach simplifies the overall system architecture while maintaining high identification accuracy
Data Source
AI summary
A device can identify a tracking device; identify, in an image or video and using a computer vision technique, a tracked asset or an identifier associated with the tracked asset, wherein the tracking device is to be used to track a location of the tracked asset; determine information identifying an association between the tracking device and the tracked asset based on identifying the tracking device and identifying, in the image or video and using the computer vision technique, the tracked asset or the identifier associated with the tracked asset; determine a link between the tracking device and the tracked asset based on the information identifying the association between the tracking device and the tracked asset; and store or provide link information that identifies the link between the tracking device and the tracked asset to track the location of the tracked asset using the tracking device.