Object Tracking via Adjacent Camera Feed Elimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object tracking methods are inefficient, especially when no tracking tag is associated with the object, leading to a wide and expanding search area with low success rates, as they rely on generalized search areas and manual approaches, which become impractical over time.
Innovation Solution
A system and method that utilize cameras to identify potential objects of interest, analyze adjacent camera feeds to determine possible travel routes, and prevent analysis of downstream feeds where the object is not detected, thereby narrowing the search scope by eliminating routes without matching objects, using image recognition and camera network analysis to plot a likely route.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If generalized search areas and manual approaches are used for object tracking, then the search can cover all possible areas, but the search area expands widely over time leading to low success rates and inefficiency
Solution Approach 1:
The patent segments the search space by dividing it into multiple camera coverage zones and systematically eliminating areas where objects are not detected. Instead of searching the entire area, the system divides the search into discrete camera fields of view and progressively narrows down by eliminating segments without detections
Solution Approach 2:
The system performs preliminary actions by proactively eliminating camera feeds and geographic areas from the search before they need to be manually reviewed. By pre-processing the search space and removing impossible routes based on absence of detections, the system reduces the workload for subsequent tracking operations
2Reliability
If all camera feeds are analyzed continuously to track objects, then no potential route is missed, but the computational complexity and time required increase significantly
Solution Approach 1:
The patent extracts and eliminates camera feeds from the analysis pipeline when they are determined to be irrelevant. By removing camera feeds that do not contain objects of interest or that cover areas already searched, the system reduces computational complexity while maintaining tracking accuracy through selective continuation of analysis on relevant feeds only
Solution Approach 2:
The system applies partial action by analyzing only the necessary subset of camera feeds required to maintain tracking accuracy. Instead of exhaustively analyzing all feeds, it performs sufficient analysis on feeds that could potentially contain the object, eliminating the need for excessive analysis of feeds that cannot contribute to finding the object
3Productivity
If manual approaches are used to narrow down search areas, then human judgment can be applied, but the process becomes impractical over time and reduces productivity
Solution Approach 1:
The patent implements self-service by enabling the system to automatically narrow down search areas without human intervention. The automated elimination of camera feeds and search spaces based on detection patterns allows the system to self-manage the tracking process, maintaining high productivity and rapid object location without requiring manual review over time
Data Source
AI summary
According to a computer-implemented method, multiple potential objects of interest are identified from a camera feed of a source camera. A number of adjacent cameras are identified along possible travel routes. For each adjacent camera, it is detected whether any of the multiple potential objects of interest are identified in an associated camera feed. For each adjacent camera whose feed does not include any potential object of interest, analysis of downstream camera feeds is prevented during tracking of the multiple potential objects of interest.