Video Object Detection via Tracking Feedback Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video surveillance systems with multiple cameras face challenges in efficiently monitoring and tracking objects due to high computational requirements and the need for extensive personnel monitoring, leading to a demand for automated image evaluation.
Innovation Solution
Implementing a feedback mechanism that connects the tracking device with a model selection system, allowing for the use of more specific object models with narrower variation ranges, reducing computing power and improving detection accuracy by selecting appropriate models based on tracking and scene parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single object model covering all variations is used for detection, then the system is simpler to operate, but detection accuracy decreases and computing power requirements increase
Solution Approach 1:
The patent segments the single object model into multiple specialized detection modules, each handling specific object variations (e.g., person, automobile, animal). This segmentation improves detection accuracy by matching specific object types with appropriate models while reducing unnecessary computational load from evaluating all variations for every detection task.
Solution Approach 2:
The system dynamically selects which detection module to use based on tracking parameters and scene parameters. This dynamic adaptation allows the system to switch between different object models appropriate to the current situation, improving detection accuracy while maintaining operational simplicity through automated model selection.
2Measurement precision
If multiple specialized detection modules are used, then detection accuracy improves, but the complexity of selecting the appropriate model increases
Solution Approach 1:
The system performs self-service by automatically selecting the appropriate detection module based on tracking parameters and scene parameters. The selection process is autonomous and does not require manual intervention, thereby improving detection accuracy through specialized models while maintaining ease of operation through automated decision-making.
3Productivity
If automated image evaluation is implemented, then productivity increases, but computational load increases
Solution Approach 1:
The system changes parameters by selecting different detection modules based on tracking parameters and scene parameters. This parameter-based selection allows the system to use computationally efficient models when appropriate while maintaining high detection accuracy, thereby increasing productivity without proportionally increasing computational load.
Data Source
AI summary
The invention relates to a device, method, computer program, and a computer program product for monitoring objects, in particular for monitoring scenes of objects captured on video. An object is thereby repeatedly detected and tracked, wherein a tracking device is fed back to a device for object model selection, so that when detected repeatedly, considering tracking parameters determined when tracking the object, the tracking parameters are fed to the selection device and can be considered for detecting.