Multi-Camera Video Stream Alteration for Obscured Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying objects of interest in video streams is challenging due to conditions such as poor lighting, black and white video, or obscured items, which makes it difficult for operators to recognize specific objects.
Innovation Solution
A computing device analyzes multiple video streams from non-overlapping cameras to identify visual traits of an object-of-interest (OOI) in one stream and alters the first video stream to visually indicate the OOI, using traits identified in a second stream.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If video streams are displayed as received from cameras, then the system complexity is low, but the ability to identify objects of interest deteriorates when lighting conditions or video quality are poor
Solution Approach 1:
A computing device is introduced as an intermediary between the camera and the display screen. This intermediary receives video streams from multiple cameras, analyzes them to identify objects of interest, and alters the video streams before displaying them. The intermediary handles the complexity of multi-camera coordination and object identification, allowing the display system to remain relatively simple while significantly improving object identification reliability.
Solution Approach 2:
The computing device performs preliminary analysis of video streams from multiple cameras before the video is displayed. By pre-identifying objects of interest and determining which cameras provide the best views, the system prepares enhanced video content in advance, ensuring that when the video is displayed, objects of interest are already highlighted and identifiable, thus improving reliability without requiring complex real-time processing at the display end.
2Reliability
If operators manually search through video streams repeatedly, then the identification accuracy improves, but the time consumption and processing resources increase
Solution Approach 1:
The computing device performs self-service by automatically analyzing video streams from multiple cameras, identifying objects of interest, and altering the video streams to highlight these objects. This automation eliminates the need for operators to manually search through video streams repeatedly, significantly reducing time loss while maintaining high identification accuracy through computational analysis.
Solution Approach 2:
The patent replaces the mechanical process of manual video review with an automated computing system. Instead of operators visually scanning through video streams, the computing device uses image processing algorithms to automatically detect and highlight objects of interest, substituting human visual inspection with computational analysis that is both faster and more consistent.
3Reliability
If multiple video streams from different cameras are analyzed, then the ability to identify obscured objects improves, but the processing resources required increase
Solution Approach 1:
The computing device applies local quality by analyzing only the specific regions and frames from multiple camera streams that are most likely to contain objects of interest. Rather than processing every pixel of every video stream equally, the system identifies promising candidates from multiple cameras and concentrates computational resources on those specific local areas, improving the ability to recognize obscured objects while managing processing resource consumption.
Solution Approach 2:
The system employs partial action by selectively analyzing portions of multiple video streams rather than processing all available video data exhaustively. By identifying objects of interest in at least one camera stream and using that information to guide analysis of other streams, the system achieves reliable object recognition without requiring full processing of all video data, thus optimizing the balance between recognition capability and processing resource usage.
Data Source
AI summary
A device, system, and method for altering video streams to identify objects of interest is provided. A devices analyzes: media data to determine one or more visual traits of an object-of-interest (OOI); and a first video stream to determine that the visual trait(s) are not identifiable at the first video stream, the first video stream being rendered at a display screen. The device identifies at least one of the visual trait(s) at a second video stream to identify the OOI at the second video stream, the second video stream associated with the first video stream. The device identifies the OOI at the first video stream using the OOI as identified in the second video stream. The device alters the first video stream at the display screen to visually indicate one or more of: the and at least one of visual trait(s) identified at the second video stream.