Visible-Camera Fire Detection Using Flow Vectors to Reject Reflections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fire detection systems using imaging devices struggle to differentiate between real fires and reflections, leading to false alarms, and thermal cameras, while effective, are costly and may not fully distinguish between real and reflected fires due to IR radiation reflection.
Innovation Solution
A method and apparatus using visible wavelength cameras and image processing techniques to detect fire by calculating flow vectors from background feature displacements in sequential images, distinguishing real fires based on air refractive index changes caused by heat, and identifying potential fire locations through region analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If thermal cameras are used to detect fire, then fire detection accuracy is improved, but system cost increases
Solution Approach 1:
The patent replaces thermal imaging cameras with visible wavelength cameras combined with image processing algorithms. The system uses flow vector calculation from background feature displacement to detect fire, substituting expensive thermal detection technology with cheaper visible light cameras and computational methods while maintaining fire detection capability
Solution Approach 2:
The patent introduces flow vector calculation as an intermediary mechanism between visible light detection and fire identification. By analyzing the displacement of background features and calculating flow vectors, the system indirectly detects the thermal effects of fire without requiring direct thermal imaging, thus reducing cost while preserving detection accuracy
2Ease of manufacture
If visible wavelength cameras are used to detect fire, then system cost is reduced, but ability to differentiate real fire from reflected fire deteriorates
Solution Approach 1:
The patent performs preliminary analysis of background features by capturing multiple sequential image frames and calculating flow vectors before making fire differentiation decisions. This preliminary action of analyzing air flow patterns through background displacement enables the system to distinguish real fire from reflected fire using visible light cameras
Solution Approach 2:
The patent uses feedback from flow vector analysis to improve fire differentiation. By continuously monitoring background feature displacement and calculating flow vectors across multiple frames, the system receives feedback about air movement patterns that helps distinguish between real fire (which causes air flow) and reflected fire (which does not), thereby improving differentiation accuracy without increasing system cost
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Effectively differentiates between real and reflected fires without bulky optics, reducing false alarms and maintaining cost-effectiveness by using visible wavelength cameras and image processing to analyze air flow patterns.
Implementation Method 1
distinguishing real fires based on air refractive index changes caused by heat
Data Source
AI summary
Various embodiments provide systems, methods, and apparatuses for detecting fire. An apparatus may capture a first image frame from a field of view, capture a second image frame from the field of view, compare a first location of a background feature in the first image frame with a second location of the background feature in the second image frame. The apparatus may calculate a flow vector using the comparison of the first location of the background feature in the first image frame and the second location of the background feature in the second image frame and differentiate between an image of a real fire and an image of a reflected fire using the flow vector.